Increasing retention of under-represented students in STEM through affective and cognitive interventions 9780841233645, 0841233640, 9780841233652

449 74 11MB

English Pages [173] Year 2018

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Increasing retention of under-represented students in STEM through affective and cognitive interventions
 9780841233645, 0841233640, 9780841233652

Table of contents :
Content: The value of theoretical frameworks --
Border crossings : a narrative framework for interventions aimed at improving URM and first-generation college student retention in stem --
Supporting STEM students through attachment theory --
Case studies : models that improved student success --
A comprehensive model for undergraduate science education reform to better serve the underserved --
Evaluation of effects of an intervention aimed at broadening participation in STEM while conveying science content --
UWM STEM CELL : accelerating the pace to academic success --
Effective strategies to improve academic success and retention in underrepresented STEM students --
Seeking to improve retention through teaching strategies and peer tutoring --
Studio format general chemistry : a method for increasing chemistry success for students of underrepresented backgrounds --
Applying innovations in teaching to general chemistry.

Citation preview

Increasing Retention of Under-Represented Students in STEM through Affective and Cognitive Interventions

ACS SYMPOSIUM SERIES 1301

Increasing Retention of Under-Represented Students in STEM through Affective and Cognitive Interventions Tara L. S. Kishbaugh, Editor Eastern Mennonite University Harrisonburg, Virginia

Stephen G. Cessna, Editor Eastern Mennonite University Harrisonburg, Virginia

Sponsored by the ACS Division of Chemical Education

American Chemical Society, Washington, DC Distributed in print by Oxford University Press

Library of Congress Cataloging-in-Publication Data Names: Kishbaugh, Tara (Tara L. S.), editor. | Cessna, Stephen (Stephen G.), editor. | American Chemical Society. Division of Chemical Education, sponsoring body. Title: Increasing retention of under-represented students in STEM through affective and cognitive interventions / Tara L.S. Kishbaugh, editor, Stephen G. Cessna, editor. Description: Washington, DC : American Chemical Society, [2018] | Series: ACS symposium series ; 1301 | "Sponsored by the ACS Division of Chemical Education." | Includes bibliographical references and index. Identifiers: LCCN 2018036994 (print) | LCCN 2018051226 (ebook) | ISBN 9780841233645 (ebook) | ISBN 9780841233652 (alk. paper) Subjects: LCSH: Science--Study and teaching (Higher)--United States. | Technology--Study and teaching (Higher)--United States. | Engineering--Study and teaching (Higher)--United States. | Mathematics--Study and teaching (Higher)--United States. | Minority college students--United States. | Dropouts--United States--Prevention. Classification: LCC Q183.3.A1 (ebook) | LCC Q183.3.A1 I52 2018 (print) | DDC 507.1/173--dc23 LC record available at https://lccn.loc.gov/2018036994

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48n1984. Copyright © 2018 American Chemical Society Distributed in print by Oxford University Press All Rights Reserved. Reprographic copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Act is allowed for internal use only, provided that a per-chapter fee of $40.25 plus $0.75 per page is paid to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. Republication or reproduction for sale of pages in this book is permitted only under license from ACS. Direct these and other permission requests to ACS Copyright Office, Publications Division, 1155 16th Street, N.W., Washington, DC 20036. The citation of trade names and/or names of manufacturers in this publication is not to be construed as an endorsement or as approval by ACS of the commercial products or services referenced herein; nor should the mere reference herein to any drawing, specification, chemical process, or other data be regarded as a license or as a conveyance of any right or permission to the holder, reader, or any other person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted work that may in any way be related thereto. Registered names, trademarks, etc., used in this publication, even without specific indication thereof, are not to be considered unprotected by law. PRINTED IN THE UNITED STATES OF AMERICA

Foreword The ACS Symposium Series was first published in 1974 to provide a mechanism for publishing symposia quickly in book form. The purpose of the series is to publish timely, comprehensive books developed from the ACS sponsored symposia based on current scientific research. Occasionally, books are developed from symposia sponsored by other organizations when the topic is of keen interest to the chemistry audience. Before agreeing to publish a book, the proposed table of contents is reviewed for appropriate and comprehensive coverage and for interest to the audience. Some papers may be excluded to better focus the book; others may be added to provide comprehensiveness. When appropriate, overview or introductory chapters are added. Drafts of chapters are peer-reviewed prior to final acceptance or rejection, and manuscripts are prepared in camera-ready format. As a rule, only original research papers and original review papers are included in the volumes. Verbatim reproductions of previous published papers are not accepted.

ACS Books Department

Contents Preface .............................................................................................................................. ix

The Value of Theoretical Frameworks 1.

Border Crossings: A Narrative Framework for Interventions Aimed at Improving URM and First-Generation College Student Retention in STEM .... 3 Stephen Cessna, Lori Leaman, and Lori Britt

2.

Supporting STEM Students through Attachment Theory ................................. 17 Jeffrey J. Carew and Brandon M. Fetterly

Case Studies: Models that Improved Student Success 3.

A Comprehensive Model for Undergraduate Science Education Reform To Better Serve the Underserved ............................................................................... 31 Jim E. Swartz and Leslie A. Gregg-Jolly

4.

Evaluation of Effects of an Intervention Aimed at Broadening Participation in STEM while Conveying Science Content ........................................................ 59 Heather Perkins, Mary Wyer, and Jeffrey N. Schinske

5.

UWM STEM CELL – Accelerating the Pace to Academic Success .................. 83 Anja Blecking

6.

Effective Strategies To Improve Academic Success and Retention in Underrepresented STEM Students ...................................................................... 97 Pamela K. Kerrigan, Ana C. Ribeiro, and Patricia A. Grove

7.

Seeking To Improve Retention through Teaching Strategies and Peer Tutoring ................................................................................................................. 109 Tara L. Kishbaugh, Steve Cessna, Lori Leaman, and Daniel Showalter

8.

Studio Format General Chemistry: A Method for Increasing Chemistry Success for Students of Underrepresented Backgrounds ................................. 131 Jane Brock Greco

9.

Applying Innovations in Teaching to General Chemistry ................................ 145 W. Christopher Hollinsed

Editors’ Biographies .................................................................................................... 153

vii

Indexes Author Index ................................................................................................................ 157 Subject Index ................................................................................................................ 159

viii

Preface While there is a growing demand for well-trained STEM graduates, retention of STEM majors is often lower than for other undergraduate majors. Additionally, populations traditionally under-represented in STEM fields, such as women, firstgeneration college students, and certain racial and ethnic populations, tend to be retained in STEM majors in lower numbers. Thus, working to improve retention of under-represented minorities (URM) and first-generation STEM majors is deemed a national STEM education priority. As a result, many grant solicitations, such as those from the NSF, call for research regarding how best to broaden participation of under-represented groups in STEM. Students who leave undergraduate STEM major programs generally do so in the first or second year, and the reason seems to be a complicated mixture of difficulty and lack of support/scaffolding with required, first- and second-year, large enrollment science and math courses, and/or an unwelcoming atmosphere at our universities, and particularly in those courses. Thus, interventions that succeed in increasing retention tend to address more than one of the reasons that students leave STEM. A symposium entitled “Increasing Retention of Under-Represented Students in Chemistry” was organized at the Fall 2017 National Meeting of the American Chemical Society, in Washington, D.C. This symposium grew out of the desire of the organizers to provide a venue for reports on successful NSF-funded initiatives such as those from solicitations for Institutional Undergraduate STEM Education (IUSE) and Scholarships for STEM majors (S-STEM), projects that seek to broaden participation of URM and first-generation students. This session was well-attended and received, and featured presentations from a variety of types of institutions, including two- and four- year schools, public and private institutions, Historically Black Colleges and Universities (HBCUs) and those that are primarily White. The audience was enthusiastic about the relevance and the timeliness of such a symposium. Unfortunately not all symposium speakers were able to contribute to this book. In particular, we note that Dr. Isiah M. Warner has already published on some aspects of Louisiana State University’s highly successful Hierarchical Mentoring Model (1). Carolyn Schick’s work using tutors in a two-year college setting (Montgomery College, MD) was also a valuable addition to the symposium, and is also published elsewhere (2). The first section of the book describes some of the underlying theories for understanding how and why students are or are not retained in science. As this field is quite broad, this section does not provide a comprehensive overview of all possible theories used to understand this problem, but instead it seeks to provide a rationale for the use of a specific theory to ground the nature of the interventions used to address poor retention of STEM majors. In addition, this section addresses ix

some of the nuances that are important to consider when applying a theoretical framework at your institution. In the second section, these theories are elucidated by case studies from diverse institutions, which have implemented strategies to address the problem of poor retention at their schools. In addition to diversity of institutions, the projects themselves are at different stages. Some, such as Swartz and Gregg-Jolly’s chapter, describe a project that is very well established. While others, such as Kishbaugh’s project, are in early stages. Some of these interventions are aimed at generating and testing better pedagogical tools and supports for a wider audience of students, such as Greco’s description of the studio model for general chemistry and Hollinsed’s approach to testing in general chemistry. Others use peer tutors, peer mentors, and/or diversity-responsive teaching materials. Others describe mentoring, intrusive advising, or cohort models. All S-STEM projects involve cohort building; examples can be found in Kerrigan’s chapter or Blecking’s chapter. Perkins’s project also describes the importance of addressing stereotypes in science. These case studies demonstrate a balance of attention towards affective dimensions, such as self-efficacy and identity, with evidence-based teaching methods. Overall this book seeks to provide other faculty and administrators with examples of effective strategies that address the many and interacting reasons why students might leave the sciences. The diversity of the institutions represented in these case studies indicates the generalizability of the issues and potential solutions. We hope that departments, chairs, deans, and other academic leaders will see how aspects of the interventions described here would fit at their universities. We are grateful to the authors for sharing their stories in the form of contributed chapters and to the many reviewers who carefully read and critiqued the work to improve its quality and clarity. Finally we gratefully acknowledge the clear guidance and support of the staff at the ACS Books Editorial Office, especially Kelsey Barham, Victoria Balque-Burns, and Amanda Koenig.

References 1.

2.

Wilson, Z. S.; Holmes, L.; deGravelles, K.; Sylvain, M. R.; Batiste, L.; Johnson, M.; McGuire, S. Y.; Pang, S. S.; Warner, I. M. Hierarchical Mentoring: A Transformative Strategy for Improving Diversity and Retention in Undergraduate STEM Disciplines. J. Sci. Educ. Technol. 2012, 21, 148–156. Schick, C. P. Trying on Teaching: Transforming STEM Classrooms with a Learning Assistant Program. In Strategies Promoting Success of Two-Year College Students; ACS Symposium Series 1280; Anna, L. J., Higgins, T., Palmer, A. M., Owens, K., Eds.; American Chemical Society: Washington, DC, 2018.

x

Tara L. S. Kishbaugh Professor of Chemistry Department Chair, Biology and Chemistry Eastern Mennonite University 1200 Park Rd Harrisonburg, Virginia 22850, United States [email protected] (e-mail) 540-432-4665 (telephone)

Stephen G. Cessna Professor of Biochemistry Daniel B. Suter Endowed Chair Eastern Mennonite University 1200 Park Rd Harrisonburg, Virginia 22850, United States [email protected] (e-mail) 540-432-4403 (telephone)

xi

The Value of Theoretical Frameworks

Chapter 1

Border Crossings: A Narrative Framework for Interventions Aimed at Improving URM and First-Generation College Student Retention in STEM Stephen Cessna,*,1 Lori Leaman,2 and Lori Britt3 1Department

of Chemistry, Eastern Mennonite University, Harrisonburg, Virginia 22802, United States 2Department of Education Eastern Mennonite University, Harrisonburg, Virginia 22802, United States 3School of Communication Studies, James Madison University, Harrisonburg, Virginia 22807, United States *E-mail: [email protected]

This volume contains several useful reports of “what works”: descriptions of specific interventions for improving STEM major retention of under-represented minority (URM) groups. In this chapter we focus on what is known about why these interventions work based on three major considerations of the lived URM and first-generation student experience in college STEM programs: preparation for rigorous college coursework, student identity and self-efficacy with science, and cultural congruence with the institution and department. Rather than describing these from a theoretical framework, these ideas are discussed using a narrative framework of border crossings. We hope this framework provides a useful means of planning and assessing interventions aimed at improving persistence for under-represented groups.

What Works and Why? The problem of low persistence and graduation rates for under-represented minority groups (URM) in STEM majors has received much attention in the past © 2018 American Chemical Society

few years in the recommendations of numerous ‘blue-ribbon’ panels and federal grant calls (1–4). The issue of first-generation college student retention has also garnered significant attention; students’ whose parents did not complete college are statistically less likely to complete college, even after controlling for academic preparation (5–9). URM students are more likely to also be first-generation college students, and thus, considerations to improve retention of either group should consider issues of both (7, 8). Numerous useful articles have appeared on these topics, including several in this volume, describing attempts at addressing this national economic and social justice problem in different institutional contexts. From these efforts, we are gaining a clearer idea of multiple effective means of improving URM and first-generation retention in STEM majors. Several recent articles summarize “what works”, including references (10) and (11). Readers searching for ways to address the issue in their own college context will hopefully find practical examples in this literature that could be readily adopted. Unfortunately, however, readers might get instead a sense that the interventions presented are only feasible in the college context from which they were written but not in their own, due to the particularities and peculiarities of their own institutional context. Such a reader might respond to these chapters by saying: “That’s great! But it would never work here.” Another means of advancing the adoption of these and other intervention strategies is by naming and clarifying why URM and first-generation STEM major retention is lower than we’d like, and then perhaps seeking to understand why and how these many interventions work. We argue that these interventions are effective and can be generalized to multiple settings, not so much because of the specifics of their design, which may not be transferable to every institutional context, but because they all hold to one or more common underlying aspects that explains the URM and first-generation experience in college STEM programs. Some Advantages and Disadvantages to Using Theoretical Frameworks In science a vast theoretical framework undergirds our research, from the broadest general theories like atomic theory and Lewis acid/base theory, to the more arcane theories and models of our sub-disciplines, or even the ‘pet theories’, hypotheses, or testable models specific to our research teams. Those of us trained in the sciences are comfortable working explicitly with and around theories in framing our research proposals, explaining our findings, and reporting our successes. Just like in scientific research, working with and from a theoretical position is critical to the effective design, assessment, and reporting on educational research and interventions. A theoretical framework for educational research and intervention design can do all of the following (adapted from Bodner and Orgill (12)): •



Better connect the current intervention or research question with the existing research knowledge base, both drawing from and potentially adding to that existing knowledge; Provide a testable general model from which to design further specific interventions and ongoing research questions; 4





More clearly articulate the why and the how for the successes or failures of the intervention (a chemist might call this identifying the “mechanism” for the intervention, although that word is ill-advised when working with humans instead of chemicals); and thereby Allow the reader to see the generalizability of the intervention to any and all institutional contexts, or to more clearly frame the limits of the generalizability (in the same way that a robust mechanistic theory applies to a wide range of chemical reaction settings yet still has limits, so might a robust theory of learning apply to a wide range of student demographic and institutional contexts, also within some limits).

Therefore, several recent solicitations from the National Science Foundation’s Directorate for Education and Human Resources ask explicitly for education theories to inform requests. For example, the 2017 IUSE solicitation, (Improving Undergraduate STEM Education, EHR-DUE, solicitation number 17-590), includes the following statement: “NSF expects that investments within the IUSE portfolio will integrate theories and findings from education research with developments at the frontiers of science and engineering research.” Additionally many of the journals that would be appropriate outlets for reports on interventions to improve URM STEM retention require a theoretical framework section to the manuscript. All told, grounding our work in a theoretical position is increasingly important, both for the success of the project, and for its dissemination. Unfortunately, however, reading about theories in education, with unfamiliar acronyms and social science terminology, is a challenge for those of us who are not trained in graduate education programs. Educational theory is foreign territory to us in much the same way that science and the academy are foreign to many of our students. Additionally, there are numerous different theoretical frameworks in the recent research on college URM and/or first-generation STEM retention, and selecting among them is a challenge. Some of these are more accessible to the non-expert than others, and some are more robust, developed, and tested (13, 14). One other concern in sorting through educational theoretical frameworks is finding something grounded in the whole student experience rather than an abstract idea: our view is that work on the URM retention problem needs a theoretical footing that starts from the perspective of the students’ life, in and out of the classroom. Lemke, in his 2001 description of socio-cultural perspectives in science education, notes the following: “Student interest in, attitudes toward, and motivation toward science, and student willingness to entertain particular conceptual accounts of phenomena depend on community beliefs, acceptable identities, and the consequences for a student’s life outside the classroom (and inside it) of how they respond to our well-intentioned but often uninformed efforts at directing their learning: uninformed insofar as we do not take into account that learning is not just a matter of whether we can understand a scientific account, but also of whether our social and cultural options in life make it in our interest to do so” (15). Lemke asks us to take seriously the idea that understanding how a student learns science requires a multi-dimensional and personal account of the whole student, including culture, belief, identity and community: the student’s 5

lived experience, their story (15). How much more so does the decision to stay or leave college or the STEM major depend on the student’s full story? At Eastern Mennonite University we have implemented several interventions for improving URM retention in STEM based around a narrative theme that builds from the students’ experience, using Glen Aikenhead’s idea of Border Crossing as a theoretical and conceptual framing analogy (16) (also see the chapter by Tara Kishbaugh below). We’ll first describe this narrative lens, and then name and discuss three components. These components are then used to explain several successful strategies. A Narrative Framework: Border Crossing Glen Aikenhead sought to offer “an account of students’ lived experiences in a science classroom by considering those experiences in terms of students crossing cultural borders, from the subcultures of their peers and family into the subcultures of science and school science” (16). In this sense, the URM and first-generation STEM college student experience can be thought of as a journey in which they must navigate crossing into the foreign culture of our mostly White university, then traverse a second crossing into the ‘micro-culture’ of first-year large enrollment STEM classrooms, while also crossing into a culture of a reductionist and materialist scientific worldview (16, 17). These new cultures can be foreign, if not directly at odds with their home culture, norms, behaviors and values. These successive cultural border crossings can be easy for some learners to navigate, but much more challenging for others (17, 18). There are three major potential challenges that URM and first-generation students face in crossing these successive borders into our STEM majors, which can decide their retention vs. attrition decisions: preparation for college, student identity and self-efficacy as both science students and scientists, and cultural congruence with the institutional climate. Here we describe each of these challenges from the perspective of the border crossing analogy in concert with some recent empirical research. We suggest that effective intervention on behalf of these students requires careful consideration of all three.

Considerations of the URM Experience Preparation for College “To me, as a freshman starting in a big auditorium type thing… It’s pretty intimidating, but cool. None of my family went to college. I was the first one. So, I felt like I was in a movie or something. So, in [my first year chemistry] class, the professor would say: ‘You remember this from high school’, like structures, polar, non-polar, and stuff like that. But I was just like ‘No, I don’t!’ I was a good student in high school, but my high school chemistry was just so bad.” [A junior biology major] In our border crossing analogy, students can enter our college STEM classrooms underprepared for the journey, and thereby be more likely to struggle 6

to find a footing in their new context. Numerous research articles note that academic preparation is a strong predictor of retention in STEM. For example, Chang and colleagues note that in a sample of more than 1500 URM students in STEM majors across the U.S., every 100-point increase in combined SAT score increased the likelihood of retention/completion in STEM majors by nearly 7% (11). Even more sobering is the 2017 SAT results reported by the College Board, demonstrating more than a 150-point average achievement gap between White or Asian students relative to URM groups. These issues go beyond race. First-generation college students also have significantly lower SAT scores, irrespective of race (8). Also, the urban or rural vs. suburban nature and the socio-economic status of the community the student come from can all have a significant impact on student college academic readiness, regardless of race (19). For example, the quotation above is from a white student from a rural U.S. high school, who also felt very unprepared for college chemistry content. There are also disparities in preparation for college that cannot be measured by the SAT. Much of college readiness is in cognitive and “soft skills” rather than content knowledge. Conley has noted four indicators of college readiness including: 1) knowledge and understanding of specific content, (arguably, that which is measured on the SAT and other high-stakes exams), 2) general academic skills such as critical thinking and problem solving skills, and specific academic skills including writing and numerical reasoning, 3) “college knowledge” – understanding the processes and cultures of college, and 4) academic behaviors and cognitive dispositions, such as self-awareness and self-monitoring (20). It is easy to see how these last two indicators might be more accessible to students whose parents and other family members went to college. Conley’s last category involves aspects of self-knowledge; other authors would add self-regulation and self-efficacy to that list (21, 22). This broader sense of student preparation is social and cognitive, and ultimately involves student identity. Student Identity “Having that [early undergraduate research] experience was really helpful, and it really made me feel like, ‘Oh! I can do science!’” [A freshman chemistry major] In our border crossing analogy, students in a science degree program must become acculturated into the new territory of the college STEM program, the University, and the discipline, taking on the norms, values, and sense of identity with the place they’ve entered. Similarly, Lave and Wenger developed a model for learning in which students’ identity formation is at the forefront (23). Teaching and learning, in this conception, is envisioned as a ‘cognitive apprenticeship’, in which students are joining the community of practicing scientists. At first it is as a peripheral but still legitimate member of that community. But then, over time and with direction from the ‘journeymen’ and ‘masters’ of that community, they become further socialized into the community of practice, taking on increasingly important tasks within the community, which garner their learning of the social mores of the community, its language and shared assumptions. The student gains 7

an identity as a member of the community of practice, both in their own eyes and in those of the community (23). In this sense, for Lave and Wenger, “learning and a sense of identity are inseparable: They are aspects of the same phenomenon” (23) (p. 115). Students who succeed in science will take on a science identity: a sense of who they are, what they are capable of, what they want to do and be, all in relation to science (24, 25). Science identity formation is a socialization process, depending on the students’ full-lived experience and multiple social interactions, including home, college academic interactions, clubs, roommates, work, and beyond. At their best, undergraduate science programs assist with or reinforce student identity formation with science and as scientists (26, 27). At their worst, our programs will privilege scientific identity formation for some groups of students over others, incidentally inviting whiter, wealthier, and more male students into community of science than others (24–27). The college student who is developing a science identity that carries her towards retention and degree completion in STEM must not only come to view herself as a potential scientist, but also develop an accurate sense of her own capacity for science and math course work. Self-efficacy is the conviction that one can successfully execute the behavior required to reach a certain goal (28). A positive self-efficacy in math and science is tightly aligned with STEM major selection and persistence in STEM (22, 29–31). Science identity and self-efficacy are both directly related to the degree of support provided from multiple quarters, including home, school, inspiring role models, tutors, and extracurricular activities (27). An example theoretical framework for URM and first-generation STEM majors that pays close attention to student self-efficacy and identity is Expectancy Value Theory (32). Expectancy Value Theory starts with the psychological construct of motivation: much of student persistence vs. attrition in STEM can be explained based on how motivated different students are to continue in the major, how much of that motivation is extrinsic (based on grades) vs. intrinsic (for the pleasure of the work itself), and how much the perceived benefits outweigh the perceived costs (3, 32, 33). A recent article by Andersen and Ward (33) is an exceptional example of a study that uses the Expectancy Value Theory framework in research related to URM retention in STEM, albeit at the high school level rather than with college students. In their model, a students’ decision to persist in STEM course work and careers is driven by two factors: 1) their personal expectations for success in that endeavor (self-efficacy); and 2) the relative value (or subjective value) that they assign to the outcomes that they imagine will be available from continuing in STEM. The expectancy aspect involves the student asking, “Can I do this?”; the value portion has them asking, “If I do this, what will happen?” Negative answers to either question are directly linked to poor persistence in STEM (33, 34). 8

Expectancy Value Theory expands the value side of the equation, noting that students make decisions for retention or attrition from the STEM pipeline based on four assessments of the relative value of attaining that goal (33): a.

b. c. d.

utility value – valuing STEM coursework because its completion is understood to provide eventual professional reward and financial payoff (an extrinsic motivation); intrinsic value – valuing and enjoying STEM coursework for its own sake; identity value – valuing science or math because it is consistent with their identity: their sense of who they are, and who they want to become; cost – the time and effort required for math and science coursework, taking away from other desired activities, and the negative responses of peers or family to those goals.

Andersen and Ward found that relative value was the significant driver for student persistence in STEM (33). In their analysis, all aspects of value (utility, intrinsic, and identity) were predictive of student plans to persist in STEM. STEM identity value (the degree to which students identify with science, or feel that “science is me!”), was the strongest predictor of student plans for STEM persistence, but was much stronger among White than URM students. In this study, academically successful URM students were less likely than their white student peers to see STEM course work and eventual careers as congruent with their identity and who they want to become (33). From these findings and others like them, it is clear that there is some cultural aspect to science identity that should be further considered. Cultural Congruity “The first month hit me hard. And I think a lot of it was being so far from home. I don’t know; it was just really hard. Now I have to put myself out there and be actively looking for things to do and for ways to get involved, but that’s hard too. That was one of the main things that was hard for me to transition to, was to try to get involved on top of all the sudden having all of this college work, and it’s so much different from home, from high school.” [A freshman biochemistry major] Our own experiences visiting cultural settings outside of our own should alert us to the challenges faced by many students in our STEM courses (35). When we travel we expect to experience differences of culture, perhaps even finding the local customs quaint. However, we might also encounter unexpected clashes between our norms and values relative to those of the culture that we are visiting, or in some cases even perceive hostilities to our home culture or identity. In these cases “culture shock” is an appropriate response. Longer-term residence in a foreign culture is known to cause stressful cognitive dissonance (36), and adaptation to a new culture often involves both psychological and sociocultural adjustment in personal identity (37). 9

In much the same way, the Latina/o student who enters the general chemistry classroom on the first Monday in September may experience cultural incongruity with the largely white academic environment, such as the freshman biochemistry major quoted above (38–40). Cultural congruity – the match of one’s cultural or personal values with those of the university (38) – has been developed as a measurable aspect of URM college student retention. Because of cultural incongruity, or a sense of not belonging to the majority culture of the institution, URM students may feel isolated, unwanted, or “token”, all of which can negatively affect their retention. URM students may find it difficult to balance “home” culture and “academic” culture, particularly if there is a perception of a biased, discriminatory, or even hostile university learning environment or campus climate (39). Students’ perception of the campus racial climate is thus a strong determinant of URM cultural congruity, and thereby of college retention (41). The major facets of campus climate that impact URM student retention include the institution’s historical legacy of inclusion (e.g. history of participation in segregation in the U.S. south), the diversity of the student body, faculty, staff and administration, and the students’ perceptions of attitudes and behaviors between racial groups on campus (39, 41). Another way of naming and describing this issue is sense of belonging: “Sense of belonging is a student’s own psychological sense of social integration resulting from the intersection of academic and social realms, which are crucial to students’ transition in college” (42). Several studies have found that URM students in STEM have a lower sense of belonging than either their white colleagues in STEM, or other URM students in non-STEM majors (43). Others have found that this poor sense of belonging can be remediated with URM students, with resulting benefits for both academic performance and health outcomes (44). Using the border crossing analogy, URM and first-generation students are in need of a tour-guide, or better yet, of a ‘cultural broker’ (35): a facilitator between their home world and that of the academy, and ultimately that of science. They need someone or some group who can help make the cultural transition smooth. Additionally, if the goal is truly to improve URM retention in this foreign land of academic science, the onus must not be entirely on the student. Rather, the cultural climate of institution, the STEM program and classroom experience, must improve to better allow for URM and first-generation science student identity formation. Perhaps most telling of the impact of campus racial and cultural climate is the finding that STEM retention is significantly higher for African American students at Historically Black-Serving Institutions (HBUCs) (11), and for Latino/a students at Hispanic Serving Institutions (HSIs) (45), than for either group at predominantly White institutions.

Implications for Intervention Design and Assessment Our view is that these three challenges should be carefully and equally considered together in designing and assessing effective programs for URM and first-generation retention in college STEM: 10

• • •

disparities in academic preparation, social cognitive preparation, including self-efficacy and identity with science, and cultural congruity with the STEM major program and university.

Several common types of interventions that specifically address the issue of academic preparation include academics-focused summer bridge programs, remedial (or developmental) first year math or writing course work, and supplemental or peer-led instruction. Meta-analyses of these efforts show mixed effects on long-term retention in STEM majors (46–48). These attempts have been shown to be most successful when they pay explicit attention to self-efficacy, identity, and cultural congruity questions. For example, Stolle-McAlister’s investigation of the summer bridge program at the University of Maryland Baltimore County indicated that the experience was successful in increasing URM STEM retention because it aided formation of academic self-confidence/self-efficacy and social community among participants. These efforts increased the African-American student participants’ academic, cultural, and social capital prior to undertaking the first semester of a STEM college career (49). Another example that is worth explaining in more detail explicitly ties all three factors together: the Howard Hughes Medical Institute-supported program at Louisiana State University (50) (Dr. Warner, the PI on the LSU-HHMI project, was a speaker at our symposium who did not contribute to this volume since this work was already published). The multi-faceted LSU-HHMI project included hierarchical mentoring, undergraduate research apprenticeship, and success courses/workshops specifically targeting under-achieving URM sophomores in STEM majors. These “flexible and adaptive professional and academic development courses” (50) functioned as weekly workshops run by program staff, affiliated mentoring faculty, and the campus Center for Academic Success, on topics including: learning styles, metacognitive learning strategies, group study, navigating competitive and collaborative academic settings, social integration, and recognizing racial and academic identities. They report an improvement from less than 20% STEM major completion among Black students prior to implementation of the program, to more than 50% completion for Black students who completed the program. The authors point toward the students’ self-monitoring and studentship skills honed primarily in the research mentoring relationships with selected faculty and graduate students, and in the workshop setting, as the primary causal factors in the programs successes (50). As noted in the above example, participation in early authentic research experience has a strong impact on URM retention in STEM. Hurtado further reveals several ways in which well-planned and well-executed undergraduate research programs that specifically target URM students in the first few years of college can be more successful in promoting retention: by enhancing student identity with science and improving academic science self-efficacy, while simultaneously developing directed career goals in science (51, 52). Similar results were found by Carpi and colleagues (30). The student-research mentor 11

relationship is central to the student formation of scientific identity, and can provide a sense of belonging to the major department and/or to science (53). Tending to the cultural congruity aspect in improving URM and first-generation STEM retention at mostly White universities might feel daunting to STEM faculty, or like this task should be someone else’s job. For example, faculty might feel that the best way to improve the cultural climate for improved URM retention is for administrators to recruit and hire more diverse faculty. While these efforts are important and undoubtedly would improve URM sense of belonging in our programs, the pace at which diverse hiring is occurring remains too slow. In addition, placing this problem solely on URM faculty members without supports is unfair to them, and also likely insufficient to address the problem. Initiatives that more fully address issues of cultural congruence in STEM programs include: •

• • •

STEM faculty development, including unconscious bias training and teaching strategies for diverse learners (see chapters by Kishbaugh and Swartz in this volume); purposeful selection and grooming of URM and first-generation upperclassmen for STEM leadership, such as peer tutors for STEM courses; recognition of URM and first-generation accomplishments, in and out of the classroom (see the chapter by Perkins in this volume); enhancing student sense of belonging through promotion of social interactions within STEM department student groups, one-on-one mentoring with faculty and upper level students, and involvement in minority-based campus groups such as the Black Student Union or the Latino Student Alliance.

In this volume, the chapters by Swartz and Kishbaugh and their respective colleagues clearly demonstrate attention to this issue, while still also addressing student preparation and identity. Both show that explicit attention to all three considerations in intervention design can lead to measurable improvement.

Conclusions Here we have attempted to explain the problem of low persistence in STEM among URM and first-generation college students based on students’ lived experience, using an analogy of border crossing into STEM college classrooms and professions. Three major facets of this experience resonate from our experience with students in our programs and our on-going program assessment: filling gaps in academic preparation, promoting a self-regulating cognitive capacity to see oneself as successful in science coursework and as a scientist, and cultural congruity with the institution and the STEM department. Our view is that explicit attention to all three of these considerations, in both design and ongoing assessment, will bring far greater gains in URM and first-generation retention than would an emphasis on just one or two. 12

A Note on Student Quotations Several quotations from students are used above to provide a greater sense of the student experience. These quotations were drawn from interviews performed in April, 2016, for an on-going project at Eastern Mennonite University. Our Institutional Review Board approved the interviews. Demographic details (gender and race) for each student are generally omitted to maintain their anonymity.

Acknowledgments Funding for this work was provided by the National Science Foundation, DUE-1611713. Much thanks to colleagues at Eastern Mennonite University, past and present, including Linda Gnagey, Tara Kishbaugh, Daniel Showalter, Susannah Lepley, and Dee Wiekle; and to Carol Hurney of Colby College.

References 1.

2.

3.

4.

5.

6.

7.

8. 9.

George, Y. S.; Neale, D. S.; Van Horne, V.; Malcolm, S. M. In Pursuit of a Diverse Science, Technology, Engineering, and Mathematics Workforce; American Association for the Advancement of Science: Washington, DC, 2001; pp 1−24. Olson, S.; Riordan, D. G. Engage to Excel: Producing One Million Additional College Graduates with Degrees in Science, Technology, Engineering, and Mathematics; Report to the President; Executive Office of the President: 2012. Graham, M. J.; Frederick, J.; Byars-Winston, A.; Hunter, A.-B.; Handelsman, J. Increasing Persistence of College Students in STEM. Science 2013, 341, 1455–1456. Chen, X. STEM Attrition: College Students’ Paths into and out of STEM Fields. Statistical Analysis Report; NCES 2014-001; National Center for Education Statistics, U.S. Department of Education: Washington, DC, 2013. Ishitani, T. T. Studying Attrition and Degree Completion Behavior Among First-Generation College Students in the United States. J. High. Educ. 2006, 77, 861–885. Pascarella, E. T.; Pierson, C. T.; Wolniak, G. C.; Terenzini, P. T. First-Generation College Students: Additional Evidence on College Experiences and Outcomes. J. High. Educ. 2004, 75, 249–284. Redford, J.; Hoyer, K. M. First-Generation and Continuing-Generation College Students: A Comparison of High School and Postsecondary Experiences. Stats in Brief; NCES 2018-009; National Center for Education Statistics, U.S. Department of Education: Washington, DC, 2017. Atherton, M. C. Academic Preparedness of First-Generation College Students: Different Perspectives. J. Coll. Stud. Dev. 2014, 55, 824–829. Engle, J.; Tinto, V. Moving Beyond Access: College Success for Low-Income, First-Generation Students; Pell Institute for the Study of Opportunity in Higher Education: Washington, DC, 2008. 13

10. Byars-Winston, A.; Estrada, Y.; Howard, C. Increasing STEM retention for underrepresented students: Factors that matter; Research Brief; Center on Education and Work, University of Wisconsin at Madison: Madison, WI, 2008. 11. Chang, M. J.; Sharkness, J.; Hurtado, S.; Newman, C. B. What Matters in College for Retaining Aspiring Scientists and Engineers from Underrepresented Racial Groups. J. Res. Sci. Teach. 2014, 51, 555–580. 12. Bodner, G.; Orgill, M. Theoretical Frameworks for Research in Chemistry/ Science Education; Pearson Prentice Hall: Upper Saddle River, NJ, 2007. 13. Tierney, W. G. Models of Minority College-Going and Retention: Cultural Integrity Versus Cultural Suicide. J. Negro Educ. 1999, 68, 80–91. 14. Demetriou, C.; Schmitz-Sciborski, A. Integration, motivation, strengths and optimism: Retention theories past, present and future. In Proceedings of the 7th National Symposium on Student Retention; Hayes, R., Ed.; The University of Oklahoma: Norman, OK, November, 2011; pp 300−312. 15. Lemke, J. L. Articulating Communities: Sociocultural Perspectives on Science Education. J. Res. Sci. Teach. 2001, 38, 296–316. 16. Aikenhead, G. S. Science Education: Border Crossing into the Subculture of Science. Sci. Educ. 1996, 27, 1–52. 17. Cobern, W. W.; Aikenhead, G. S. Cultural Aspects of Learning Science. Scientific Literacy and Cultural Studies Project; National Association for Research in Science Teaching: 1997; 13. 18. Phelan, P.; Davidson, A. L.; Cao, H. T. Students’ Multiple Worlds: Negotiating the Boundaries of Family, Peer, And School Cultures. Anthropol. Educ. Q. 1991, 22, 224–250. 19. Graham, S. E.; Provost, L. E. Mathematics Achievement Gaps between Suburban Students and Their Rural and Urban Peers Increase over Time; Issue Brief; Carsey Institute: 2012, Vol. 52, pp 1−8. 20. Conley, D. T. Redefining College Readiness; Educational Policy Improvement Center: Eugene, OR, 2007. 21. Zimmerman, B. J. A Social Cognitive View of Self-Regulated Academic Learning. J. Educ. Psychol. 1989, 81, 329–339. 22. Chemers, M. M.; Hu, L.-T.; Garcia, B. F. Academic Self-Efficacy and First Year College Student Performance and Adjustment. J. Educ. Psychol. 2001, 93, 55–64. 23. Lave, J.; Wenger, E. Situated Learning: Legitimate Peripheral Participation; Cambridge University Press: Cambridge, U.K., 1991. 24. Brickhouse, N. W.; Potter, J. T. Young Women’s Scientific Identity Formation in an Urban Context. J. Res. Sci. Teach. 2001, 38, 965–980. 25. Carlone, H. B.; Johnson, A. Understanding the Science Experiences of Successful Women of Color: Science Identity as an Analytic Lens. J. Res. Sci. Teach. 2007, 44, 1187–1218. 26. Egan, M. K.; Hurtado, S.; Garibay, J.; Herrera, F. Accentuating Advantage: Developing Science Identity During College. Presented at the American Educational Research Association Conference, UCLA, Los Angeles, CA, April, 2012. 14

27. Aschbacher, P. R.; Li, E.; Roth, E. J. Is Science Me? High School Students’ Identities, Participation and Aspirations in Science, Engineering, and Medicine. J. Res. Sci. Teach. 2010, 47, 564–582. 28. Bandura, A. Self-Efficacy: Toward a Unifying Theory of Behavioral Change. Psychol. Rev. 1977, 84, 191–215. 29. Betz, N. E.; Hackett, G. The Relationship of Mathematics Self-Efficacy Expectations to The Selection Of Science-Based College Majors. J. Vocat. Behav. 1983, 23, 329–345. 30. Carpi, A.; Ronan, D. M.; Falconer, H. M.; Lents, N. H. Cultivating Minority Scientists: Undergraduate Research Increases Self‐Efficacy and Career Ambitions for Underrepresented Students In STEM. J. Res. Sci. Teach. 2017, 54, 169–194. 31. Trujillo, G.; Tanner, K. D. Considering the Role of Affect in Learning: Monitoring Students’ Self-Efficacy, Sense of Belonging, and Science Identity. CBE Life Sci. Educ. 2014, 13, 6–15. 32. Eccles, J. S.; Wigfield, A. Motivational Beliefs, Values, and Goals. Annu. Rev. Psychol. 2002, 53, 109–132. 33. Andersen, L.; Ward, T. J. Expectancy‐Value Models for the STEM Persistence Plans of Ninth‐Grade, High‐Ability Students: A Comparison Between Black, Hispanic, and White Students. Sci. Educ. 2014, 98, 216–242. 34. VanZile-Tamsen, C. The predictive power of expectancy of success and task value for college students’ self-regulated strategy use. J. Coll. Stud. Dev. 2001, 42, 233–241. 35. Aikenhead, G. S.; Jegede, O. J. Cross-Cultural Science Education: A Cognitive Explanation of a Cultural Phenomenon. J. Res. Sci. Teach. 1999, 36, 269–287. 36. Maertz, C. P.; Hassan, A.; Magnusson, P. When Learning is Not Enough: A Process Model of Expatriate Adjustment as Cultural Cognitive Dissonance Reduction. Organ. Behav. Hum. Decis. Process. 2009, 108, 66–78. 37. Berry, J. W.; Phinney, J. S.; Sam, D. L.; Vedder, P. E. Immigrant Youth in Cultural Transition: Acculturation, Identity, and Adaptation Across National Contexts; Lawrence Erlbaum Associates Publishers: Mahwah, NJ, 2006. 38. Gloria, A. M.; Kurpius, S. E. R. The Validation of the Cultural Congruity Scale and the University Environment Scale with Chicano/A Students. Hisp. J. Behav. Sci. 1996, 18, 533–549. 39. Cole, D.; Espinoza, A. Examining the Academic Success of Latino Students in Science Technology Engineering and Mathematics (STEM) Majors. J. Coll. Stud. Dev. 2008, 49, 285–300. 40. Gloria, A. M.; Hird, J. S.; Navarro, R. L. Relationships of Cultural Congruity and Perceptions of the University Environment to Help-Seeking Attitudes by Sociorace and Gender. J. Coll. Stud. Dev. 2001, 42, 545–562. 41. Hurtado, S.; Clayton-Pedersen, A. R.; Allen, W. R.; Milem, J. F. Enhancing Campus Climates for Racial/Ethnic Diversity: Educational Policy and Practice. Rev. High. Educ. 1998, 21, 279–302. 42. Hurtado, S.; Han, J. C.; Sáenz, V. B.; Espinosa, L. L.; Cabrera, N. L.; Cerna, O. S. Predicting Transition and Adjustment to College: Biomedical 15

43.

44.

45.

46. 47.

48.

49.

50.

51.

52.

53.

and Behavioral Science Aspirants’ and Minority Students’ First Year of College. Res. High. Educ. 2007, 48, 841–887. Johnson, D. R. Campus Racial Climate Perceptions and Overall Sense of Belonging Among Racially Diverse Women in STEM Majors. J. Coll. Stud. Dev. 2012, 53, 336–346. Walton, G. M.; Cohen, G. L. A Brief Social-Belonging Intervention Improves Academic and Health Outcomes of Minority Students. Science 2011, 331, 1447–1451. Crisp, G.; Nora, A.; Taggart, A. Student Characteristics, Pre-College, College, and Environmental Factors as Predictors of Majoring in and Earning a STEM Degree: An Analysis of Students Attending a Hispanic Serving Institution. Am. Educ. Res. J. 2009, 46, 924–942. Sablan, J. R. The Challenge of Summer Bridge Programs. Am. Behav. Sci. 2014, 58, 1035–1050. Barnett, E. A.; Bork, R. H.; Mayer, A. K.; Pretlow, J.; Wathington, H. D.; Weiss, M. J. Bridging the Gap: An Impact Study of Eight Developmental Summer Bridge Programs in Texas; National Center for Postsecondary Research: New York, NY, 2012. Springer, L.; Stanne, M. E.; Donovan, S. S. Effects of Small-Group Learning on Undergraduates in Science, Mathematics, Engineering, and Technology: A Meta-Analysis. Rev. Educ. Res. 1999, 69, 21–51. Stolle-McAllister, K. The Case for Summer Bridge: Building Social and Cultural Capital for Talented Black STEM Students. Sci. Educator 2011, 20, 12–22. Wilson, Z. S.; Holmes, L.; Sylvain, M. R.; Batiste, L.; Johnson, M.; McGuire, S. Y.; Pang, S. S.; Warner, I. M. Hierarchical Mentoring: A Transformative Strategy for Improving Diversity and Retention in Undergraduate STEM Disciplines. J. Sci. Educ. Techol. 2012, 21, 148–156. Hurtado, S.; Cabrera, N. L.; Lin, M. H.; Arellano, L.; Espinosa, L. L. Diversifying Science: Underrepresented Student Experiences in Structured Research Programs. Res. High. Educ. 2009, 50, 189–214. Corwin, L. A.; Graham, M. J.; Dolan, E. L. Modeling Course-Based Undergraduate Research Experiences: An Agenda for Future Research and Evaluation. CBE Life Sci. Educ. 2015, 14, 1–13. Thiry, H.; Laursen, S. L. The Role of Student-Advisor Interactions n Apprenticing Undergraduate Researchers into a Scientific Community of Practice. J. Sci. Educ. Technol. 2011, 20, 771–784.

16

Chapter 2

Supporting STEM Students through Attachment Theory Jeffrey J. Carew1 and Brandon M. Fetterly2,* 1Department

of Education and Psychology, University of Wisconsin-Fox Valley, 1478 Midway Road, Menasha, Wisconsin 54962, United States 2Department of Chemistry, University of Wisconsin-Richland, 1200 US HWY 14 W, Richland Center, Wisconsin 53581, United States *E-mail: [email protected]

Attachment theory is concerned with how childrens’ early relationships affect their development and their capacity to form later relationships. The inability to form stable relationships later in life can undermine the transition to college. Attachment theory may provide understanding about what students bring into the new college environment in college, as well as how one can utilize the past to modify unhelpful expectations and failed strategies that may aid in the development of new, positive relationships. The ability of students to become attached and identify themselves as a member of the scientific community can help to support them as they navigate their educational experience.

STEM Education and Persistence For many decades, STEM educators have been concerned with persistence issues and high drop rates. Students lose interest in coursework and switch to other fields of study. A negative instructional experience in a class prevents students from continuing on to other classes within the discipline. An introductory class in another non-STEM discipline may then prove more interesting and capture the educational fire of the student. The potential reward of employment or income in a STEM field does not justify the apparent effort required to continue through © 2018 American Chemical Society

graduate studies. Every instructor in a STEM discipline could recount story upon story of talented students who chose not to continue. Anecdotally, it appears as if intelligence and talent have much less of an influence on success in STEM fields than does the sheer will to succeed. The rejection of these majors serves just as much to be a rejection of the instructors and role models of the STEM community as it does of the disciplines themselves. The reputation of the unapproachable college science instructor is all to often reinforced among the students. Popular media carries the notion that science is hard and instructors delight in weeding out the weak, culling the herd of the incapapble, a notion that scientific professionals are ill-equipped to dissuade. These observations have not been lost on the scientific community. Thirty years ago, the National Academy of Science recognized the significant attrition rates in STEM fields (1). The losses are quite sobering. At that time it was noted that over half of STEM majors were lost to other majors in the first year. Furthermore, only 35% persisted to graduation and successfully earned a baccalaureate degree. It was also noted (2) that students entering STEM majors were significantly more academically able than their non-STEM peers. It is not surprising, then, that the proportion of students entering STEM majors is rather low. Students are exiting STEM majors at alarming rates in spite of the fact that these students are intellectually able and enter college with genuine motivation to study in the field. The National Association of State Universities and Land Grant Colleges reported in 1969 that the United States led the world in production of baccalaureate engineers. Just twenty years later, Japan outpaced the United States in this metric, even though its population is half that of the US. To complicate matters, evidence suggests that secondary school science curricula serve to make the job of college STEM education all the more difficult (3). Students leave high school with misconceptions about what science is and scientific professionals do. It is no surprise then that disillusioned students choose to leave STEM majors with the realization that their major is not turning out to be what they expected. Collegiate STEM education is not free from blame either. STEM disciplines have a long standing reputation for being disproportionately white and male in its makeup (4). Moreover, as efforts to increase gender and racial diversity were making progress, an influx of non-native teaching assistants and faculty were shown to be a new barrier retention of STEM students (5). Although the language barrier is an obvious reason for this, it is not the only one. Foreign instructors coming from widely different social backgrounds than those in the U.S. may bring their own cultural norms with them. Subtle unconscious biases can even exist in environments where so much work has been done to eradicate them (6). These issues can serve as yet another barrier against retention of underrepresented minorities and women in particular. Talented individuals are being lost at a time when workforce needs are changing rapidly. Although underrepresented minorities have a strong desire to achieve these degrees, African-American STEM students leave their chosen major twice as often as white students (7). Simply improving access to higher education is not enough. These students, as do all, need access to academic and social support to aid in the successful transition to college coursework and life in general. 18

Yet another variable in the retention of STEM students are the students themselves. In order for an educational institution to be able to retain a student, that student must want to persist though difficult classes and stressful moments to achieve the reward of the degree. The motivation levels of students and their self-confidence directly relate to the effort expended to persist (8). In fact, motivational patterns of students directly shape the responses to success and failure, strongly influencing their performance (9). Although it would be tempting and easy to continue to list the shortcomings of STEM education, doing so only further identifies a long-standing problem and does not make any attempts to identify potential solutions. The purpose of this monograph is to identify the efforts of pedagogy, structure, and culture that serve to aid in retention of students in STEM majors. To that end, all these efforts hold one significant commonality, namely, to attach students to faculty and peers in their chosen major. That attachment serves to help students have a sense of belonging and identify as a member of the scientific community. Furthermore, it gives STEM students a secure place from which to reach out and explore, to grow and achieve beyond what they thought they were capable of alone. It is the existence of these attachments and the level to which students bond with faculty and peers that is of concern in this chapter.

History of Attachment Theory The notion that children “become” their parents has likely been around since the beginning of known history. The explanations of why and how we are the way we are in adolescence and adulthood (and why we either take after or not take after our parents) have only been qualified and quantified within the most recent 50 years. This formalizing took many forms, but at issue in this study is Attachment theory and its central concepts of “attachment” and “attachment style.” Traditionally, attachment describes the tendancy of a child to seek comfort and protection from a caregiver, as well as the caregiver providing this attention. Attachment theory proposes that “attachment” is the connection, closeness, and security humans experience in infancy/childhood with our primary caregiver(s), and that “attachment style” is the degree of this connection, closeness, and security. Whether this connection between infant/child and caregiver exists at all, and to what degree, is said to determine our level of security in adolescence and adulthood. Children need a relationship with a sensitive and responsive caregiver to form the basis of a secure attachment. By the age of nine months, children begin to avoid strangers and gravitate to a small number of people to whom they are attached. These attachments are formed regardless of the environment of the child, but insecure or disorganized attachments result from dysfunctional environments. Initially, children need close physical contact to build and maintain attachments, but as a child ages, attachments can be sustained over time and space with the aid of constant communication. Developing proper secure attachment early on is a key indicator for how relationships will be formed in the future. Developing appropriate relationships early is related to subsequent 19

later persistence and adaptation in new social situations. Secure attachment is important because it leads not only to more empathy and higher self-esteem, but also better stress management as a person ages. Children at risk for unhealthy attachments, children placed in out of home foster care for example (10), require special efforts to develop secure attachments to adult caregivers. Attachment theory also posits that the attachment style forged in infancy/ childhood follows us into adulthood (11). If an infant or child enjoyed a happy and safe upbringing, then this same safety and security would also be enjoyed as an adult. This becomes troubling when that same infant or child did not have a positive upbringing. Attachment theory suggests that if the aforementioned closeness and security is interfered with then the child will retain those negative feelings and be unable to form emotional bonds in adulthood. Attachment theory originated when a developmental psychologist, John Bowlby, suggested that infants have a “secure base,” or a person to turn to for unwavering safety and comfort (12). This secure base allowed a normal, safe, emotionally sheltered infancy/childhood. Bowlby emphasized that this feeling of safety led to a feeling of closeness with that person, known as “attachment.” The word “attachment” in this context refers to the connection an infant/child establishes with his/her caregiver. This caregiver is usually a parent but can take on the forms of teacher, social worker, or other trusted adult and/or authority figure. The degree or level of that close and secure connection the infant/child forges with this caregiver is called “attachment style.” Bowlby also asserted that infant/child to caregiver attachment was an “all or nothing” process. Shortly thereafter, one of Bowlby’s students, Mary Ainsworth, showed the process of an infant/child attaching to a caregiver was not an “all or nothing” process. To demonstrate this phenomenon, she developed the Strange Situation test (12). Now used to test fathers, grandparents, and even pets, it originally gauged infants’ reactions when their mothers exited and entered a room. Correlations between infants’ reactions and known characteristics revealed potential attachment styles between infant/child and mother. Ainsworth’s work (13) led to the conclusion that babies have an innate biological requirement to seek proximity to and protection from an adult for their own survival. This drive for closeness, if the attachment is to a protective and loving adult, promotes secure attachment, helping children feel safe. Attachment is the special lasting bond a child has formed with at least one adult. Specifically, it is the sense of safety and security a child has when in the company of that particular adult (14). The child learns what behaviors attract the attention of the adult to provide positive feelings in the child. The infant’s/child’s feelings become so associated with the adults who provide them that the young child quickly develops a selective attachment to the adult (15). The physically and emotionally available adult receives pleasure from spending time with the child as well. A positive attachment creates a positive feedback loop between both the parent and the child. This positivity is reflected in future relationships with others. A negative attachment style fosters just the opposite (16). Attachment development continues as a child ages. Toddlers will begin to extend their attachments to siblings and other close adults. When at school age, children who are securely attached develop the ability to understand the feelings 20

of others. At this time they develop the ability to manage themselves to work with others (15). As a child progresses to adolescence, secure attachment leads to fewer risky behaviors because parents encourage autonomy within clear limits in a warm, loving environment (17). Ainsworth’s further research would relate the style of attachment obtained in infancy/childhood to the ability to form and maintain relationships in adulthood (11). Bowlby’s original work and Ainsworth’s follow-up research laid the framework for what would later become a spin-off of Attachment theory, “Adult Attachment theory”.

Types of Attachment Originally, as a result of the Strange Situation, two attachment patterns, secure and insecure attachment were identified. Later work expanded upon insecure attachment, identifying two distinct insecure attachment types (18). These four attachment types, or attachment styles, are detailed below. A secure attachment style in infancy/childhood is marked by an infant/child who is assured of a secure base. This assurance is brought on by a caregiver who is caring, nurturing, affectionate, and quick to respond to the needs of the infant/child. The infant/child is therefore happy, independent, and well-adjusted during the early years. A secure attachment style in infancy/childhood leads to an autonomous style in adulthood, in which the adult brings forth the feelings of security and is able to be independent, socially relaxed, and comfortable with intimacy. It is thought that approximately 60% of adults in the U.S. identify with the secure attachment style. An ambivalent attachment style in infancy/childhood is marked by an infant/child who has been intermittently deprived of a secure base. This occurs when a caregiver is unreliable, disengaged from the caregiving relationship, and is sometimes present but often absent when needed. The infant/child becomes unsure of having its needs met and therefore becomes apprehensive about trusting the caregiver. This has the net effect of rendering the infant/child unresponsive, emotionally distant, and disengaged from the caregiver since it does not know when or if it will feel secure and/or have its needs met. An adult who has experienced an ambivalent attachment style should expect to be overly independent and unable to confide in or share feelings with another. Adults identifying with this attachment style will likely have difficulty with true intimacy and be given to promiscuity and casual sexual encounters that contain no emotional connection. It is thought that approximately 20% of adults in the U.S. identify with the ambivalent attachment style. An avoidant attachment style in infancy/childhood is marked by an infant/child who is insecure and anxious. The infant/child has learned to be unsure of the behavior of its caregiver, and is therefore nervous, confused, and suspicious of not only the caregiver but others as well. Avoiding others, particularly strangers, this attachment style is brought on by a caregiver who is inconsistent in attitude, behavior, and function in tending to the needs of the infant/child. An avoidant attachment style in infancy/childhood leads to an adult 21

who is needy, dependent, and often emotionally unstable in intimate relationships. An adult with this attachment style should expect his/her insecurity to create difficulties in committing to a love relationship, and the dependence and anger to produce instability if/when a relationship ends. A disorganized attachment style in infancy/childhood is marked by an infant/child who is angry, depressed, and emotionally unresponsive. The infant/child has learned to be sullen due to a caregiver who is frightening, neglectful, and abusive…the timing of which is unpredictable, further exacerbating the uncertainty and increasing the fear. A disorganized attachment style in infancy/childhood leads to an adult who is fearful and confused. An adult with this attachment style tends to avoid social situations, be unable to place trust in another, and fear intimacy with others. Having never learned these coping skills in infancy/childhood, he/she is unable to self-soothe as an adult. Without therapy, this adult will have difficulty forming and maintaining relationships, including his/her own spouse and children. The remaining 20% of adults in the U.S. (those who do not identify with either secure or ambivalent attachment styles) identify with either the latter avoidant attachment style or disorganized attachment style. Subsequent research built on this earlier work show that attachment patterns continue though later life stages (19). Four new descriptors were developed to better acknowledge the level of security within the person and in relation to others. The first scale is a measure of the amount of anxiety a person has in making attachments. The second is the level of avoidance one has in attaching to others. This leads to a scale with four quadrants where a person is ranked according to anxiety and avoidance, leading to four adult attachment types as shown in Figure 1.

Figure 1. Adult attachment styles Adults with low anxiety and avoidance are securely attached. These are people comfortable with autonomy and intimacy. They believe that they are loveable people, worthy of support. They see others around them as responsive and available to reciprocate acts of affection. In sum, they see the world as a safe and predictable place to branch out and explore who they are. 22

Adults with low avoidance and high anxiety are classified as preoccupied. Preoccupied people are just that, preoccupied with relationships and quite emotionally reactive. Individuals showing this attachment style tend to be hyper-vigilant in identifying threats. In many situations, they notice subtle emotional cues and fixate on them, even though the cues may be purely imaginary. Low anxiety and high avoidance is indicative of someone who is dismissive in their attachments. These people exhibit independence bordering on counterdependence. Often, they ignore the social cues signaling rejection. If the cue is unable to be ignored, then it is commonly dismissed. Dismissive people try to avoid feeling negative emotions or vulnerability, so closeness in relationships is similarly avoided. The final adult attachment style is seen in people who have high anxiety and high avoidance. Adults like this are described as being fearful. They have a belief that they deserve to be rejected. Because of this, they have high emotional reactivity as well. Internally, these adults desire intimacy and closeness, but become extremely uncomfortable when people get too close. This is seen by others as conflicting mixed signals. The behavior is best described as a combination of the preoccupied and dismissive attachments present simultaneously. It is possible to have more than one attachment style, or a combination of two or more. It is also possible to change one’s attachment style through Attachment Therapy (AT), which is a therapy specifically designed to “overwrite” the original attachment style. Alternatively, a method called “Earned Secure Attachment” may be used. This technique involves determining one’s own attachment style, recognizing others’ styles, and acting to counteract negative aspects of your own style and/or engaging in love relationships that introduce insecure styles or feed negative facets of your own style.

Attachment Theory and the College Transition So, how does this psychological construct relate to academics? Specifically, why might attachment style matter when it comes to enrollment in a STEM program? It matters because it appears that there is a connection between attachment style and several aspects of academia. It appears to be related to enrollment and completion, however, the literature is limited in its examination of and applicability to enrollment in specific programs. There is a strong correlation between attachment security and overall academic performance (20). In addition, secure attachment and overall psychological health in college students is strongly related (21). Insecurely attached individuals might feel that their needs cannot be met, and tend to experience more lonliness than others, possibly culminating in self-esteem and depression problems (22). Individuals with self-esteem and depression problems may turn to drug use for escape (23). Individuals who are securely attached, on the other hand, tended not to report significant alcohol use (24). As a student enters college, they are often separated from previous friends and family upon entering the new environment. Secure attachment provides the matriculating 23

student the comfort to explore the new environment while knowing that there is a secure support base available to him/her when needed. Attachment theory has evolved from the early days of Bowlby and Ainsworth to now include relational tangents such as emotional competence and development of self (25), attachment injury and forgiveness (26), self-esteem and depression (27), and anxiety and locus of control (28). The most recent literature regarding Attachment theory seems to be a multi-pronged focus on academics. Specifically, classroom behavior and self-efficacy (29), academic achievement and motivation (30), and even the very completion of education (31). There are a number of ways we as educators might best make use of Attachment theory to aid in student success. The first is by intervention strategies and the second is by using statistical data to pair attachment style with academic success, particularly in STEM. A number of intervention strategies have been attempted to aid in the transition to college and persistence of STEM students. There are three common intervention strategies used to aid in the persistence of STEM students vis-à-vis attachment. One, connecting students to expert faculty through research can be a powerful experience and learning tool.. Two, “active learning” helps to engage students with the subject matter through participation in scientific thinking and problem solving with their peers, mimicking much of what professional scientists do in their professional lives. This strategy is typically found in the classroom and has the benefit of building a community by linking peers to support them in their study. Finally, “learning communities” are support structures, either physical or virtual, that build attachment among peers outside of the classroom. The students involved have a place they can go and feel welcome to work on course content and get the help they need to persist. Active learning is defined as any activity in which every student must work on solving a problem or create knowledge. These types of classroom environments have been shown to improve STEM persistence (32). These techniques have been shown to be possible, even in large lecture courses (33). The faculty at Eastern Mennonite University have taken steps to use new teaching strategies in their classrooms. Their work can be read in greater detail in Chapter 7. Another approach that allows for more time for active learning is engaging in a “flipped classroom.” In this teaching environment, students watch didactic lectures outside of class. Since these lectures do not require much student interaction, they are easily extracted from the classroom, freeing precious class time. The in-class minutes are then repurposed to active learning environments where students work together to solve problems and actively build knowledge. The instructor in this case is more of a facilitator, intervening to aid students, but not meddling if students are working well. This type of general chemistry class is seen at Johns Hopkins University and is described in Chapter 9. Learning environments such as these allow students to rely upon each other to help build conceptual understanding. Fearful students are comforted by the fact that they’re not alone in the situation, others are there to help. Students who are preoccupied with attachments will find them easier to make in this environment, where structure and clear class rules provide safe boundaries for the external learning environment. Learning communities are spaces where students can connect with and learn from each other. Care must be taken that all students feel welcome and are 24

included in such communities. Underrepresented students are less likely to form study groups, so miss out on the key features of such communities. The Learning Assistant program at Montgomery College provides peer models who serve to help facilitate these learning groups. Learning assistants are also free to provide extra support and structured help sessions outside of class time. This intervention allows attachments to be made to a more approachable peer who works with the instructor to provide extra support and a consistent message. This program is described in greater detail in Chapter 5. Engaging students in research is an excellent way to build attachments between students and faculty members (34). Unfortunately, most undergraduate students are not given access to research experiences until the end of their baccalaureate degree, when much attrition has already taken place. In fact, students who engage in research experiences within two years are more likely to persist to earn their degree (35). The STEM-Inspire program at the University of Wisconsin-Milwaukee provides students with opportunites for participants to attach to research opportunities. The old Greek proverb, “a leopard cannot change its spots” is highly relevant to attachment style. Just as a leopard cannot change its spots, humans cannot alter their innate nature. Attachment styles are difficult to change, even with therapy. Therefore, changing attachment styles of matriculating freshman or those entering the STEM field would not be a viable option to increasing enrollment, retention, and success in STEM programs. Identifying which attachment style a student displays, however, could help to direct the student to inteventions that increase the chances of healthy attachment and success in college. A measure of this type has been shown to be insightful when examining the transition to college for women and underrepresented minorities (36). Similar effects are seen regardless of gender, racial, or ethnic heritage, with these effects being stronger when students reside away from the caregivers to which there is an attachment (37). The authors recommend the use of the Adult Attachment Scale (AAS) (38) to assess whether attachment style is a useful measure for STEM field enrollment, retention, and success. The AAS measures the first three of the previously discussed attachment styles (the scale does not assess the disorganized attachment style). The AAS is a validated, 5-point, 18-question Likert-style questionnaire that measures three dimensions: close, depend, and anxiety. The “close” dimension gauges participant comfort level with closeness and intimacy. The “depend” dimension gauges participant perception of how much he/she would be able to depend on others if the need arose. The “anxiety” dimension gauges participant concern about being unloved, unwanted, or abandoned. Specifically, it uses these three dimensions to determine whether a person has a secure, avoidant, or ambivalent (anxious) attachment style. Scoring high on the close and depend dimensions while low on the anxiety dimension indicates a “secure” attachment style. If Attachment theory is to be used to increase enrollment, retention, and success in STEM fields, we feel that it could be best used to assess one’s attachment style in conjunction with academic aptitude. This additional assessment could be given to incoming, matriculating, and transfer students as a way to gauge their attachment style. Since those with secure attachment style tend to have a higher rate of success in academia, this correlation may be useful 25

metric in encouragement or incentives for students deemed more likely to succeed in STEM fields. Identifying the attachment style a particular student displays can provide insight into which intervention strategies would be most successful in increasing the persistence rate of students in STEM majors. This has been a brief synopsis of the research, active learning, and learning community interventions. The later chapters of this monograph contain detailed examples of all three. In addition to intervention strategies, a closer look at the attachment styles of incoming students may offer some insight into overall success rates in STEM.

References 1.

2. 3. 4. 5. 6.

7.

8. 9. 10.

11. 12. 13.

14.

National Academy of Sciences. Nurturing Science and Engineering Talent: A Discussion Paper; The National Academies Press: Washington, DC, 1987; DOI: https://doi.org/10.17226/18894. Green, K. C. A profile of undergraduates in the sciences. Am. Scientist 1989, 77, 475–481. Mullis, I. V. V.; Jenkins, L. B. The Science Report Card; Educational Testing Service: Princeton, 1988. Heylin, M. Chem. Eng. News 1987, 65, 3. Vetter, B. M. Participation of women and minorities in science on decline. Science 1988, 239, 653–654. Moss-Racusin, C. A.; Dovidio, J. F.; Brescoll, V. L; Graham, M. J.; Handelsman, J. Science faculty’s subtle gender biases favor male students. Proc. Natl. Acad. Sci. U.S.A. 2012, 102, 16474–16479. National Academy of Sciences. Expanding Underrepresented Minority Participation: America’s Science and Technology Talent at the Crossroads; The National Academies Press: Washington, DC, 2011. Bandura, A. Human agency in social cognitive theory. Am. Psychol. 1989, 44, 1175–1184. Dweck, C. S. Motivational processes affecting learning. Am. Psychol. 1986, 41, 1040–1048. Smyke, A. T.; Breidenstine, A. Foster Care. In Handbook of Infant Mental Health; Zeanah, C. H., Ed.; Guilford Press: New York, NY, 2009; pp 500−515. Ainsworth, M.; Bowlby, J. An ethological approach to personality development. Am. Psychol. 1991, 46, 333–341. Ainsworth, M. Attachments beyond infancy. Am. Psychol. 1989, 44, 709–716. Bretherton, I. The origins of attachment: John Bowlby and Mary Ainsworth. In Attachment Theory: Social, Develpmental, and Clinical Perspectives; Goldberg, S., Muir, R., Kerr, J., Ed.; Analytic Press: Hillsdale, NJ, 1995; pp 45−84. Attachment: What Works? http://csefel.vanderbilt.edu/briefs/wwb_24.pdf (accessed March 4, 2018). 26

15. Schofield, G.; Beek, M. Attachment Handbook for Foster Care and Adoption; BAAF: London, UK, 2006. 16. Schofield, G.; Beek, M. The Secure Base Model: Promoting Attachment in Foster Care and Adoption; BAAF: London, UK, 2014. 17. Moretti, M. M.; Peled, M. Adolescent-parent attachment: Bonds that support healthy development. Paediatr. Child Health 2004, 9, 551–555. 18. Shemmings, D. Attachment in Children and Young People; Frontline Resources: Devon, UK, 2016. 19. Bartholomew, K. Avoidance of intimacy: an attachment perspective. J. Soc. Person. Relat. 1990, 7, 147–178. 20. Larose, S.; Bernier, A.; Tarabulsy, G. M. Attachment state of mind, learning dispositions, and academic performance during the college transition. Dev. Psychol. 2005, 41, 281–289. 21. Frey, L. L.; Beesley, D.; Miller, M. R. Relational health, attachment, and psychological distress in college women and men. Psychol. Women Q. 2006, 30, 303–311. 22. Mikulincer, M.; Shaver, P. R. Attachment in Adulthood, 2nd ed.; Guilford Press: New York, NY, 2017. 23. Vungkhanching, M.; Sher, K. J.; Jackson, K. M.; Parra, G. R. Relation of attachment style to family history of alcoholism and alcohol use disorders in early adulthood. Drug Alcohol Depend. 2004, 75, 47–53. 24. Brennan, K. A.; Shaver, P. R. Dimensions of adult attachment, affect regulation, and romantic relationship functioning. Pers. Soc. Psychol. Bull. 1995, 21, 267–283. 25. Thompson, R. Emotional competency and the development of self. Psychol. Inq. 1998, 9, 308–309. 26. Makinen, J. A.; Johnson, S. M. Resolving attachmemt injuries in couples using emotionally focused therapy: steps toward forgiveness and reconciliation. J. Consult. Clin. Psychol. 2006, 74, 1055–1064. 27. Lee, A.; Hankin, b. L. Insecure attachment, dysfunctional attitudes, and low self-esteem predicting prospective symptoms of depression and anxiety during adolescence. J. Clin. Child Adolesc. Psychol. 2009, 38, 219–231. 28. Dilmac, B.; Hamarta, E.; Arslan, C. Analysing the trait anxiety and locus of control of undergraduates in terms of attachment styles. Educ. Sci.: Theory Pract. 2009, 9, 143–159. 29. Kurland, R. M.; Siegel, H. I. Attachment and academic classroom behavior: self-efficacy and procrastination as moderators on the influence of attachment on academic success. Psychology 2016, 7, 1061–1074. 30. Learner, D. G.; Kruger, L. J. Attachment, self-concept, and academic motivation in high-school students. Am. J. Orthopsychiatry 1997, 67, 485–492. 31. Marcus, R. F.; Sanders-Reio, J. The influence of attachment on school completion. Sch. Psychol. Q. 2001, 16, 427–444. 32. Haak, D. C.; HilleRisLambers, J.; Pitre, E.; Freeman, S. Increased structure and active learning reduct the achievement gap in introductory biology. Science 2011, 332, 1213–1216. 27

33. Handlesman, J.; Miller, S.; Pfund, C. Scientific Teaching; W.H. Freeman: New York, NY, 2011. 34. Russell, S. H.; Hancock, M. P.; McCullough, J. Benefits of undergraduate research experiences. Science 2007, 316, 548–549. 35. Nagada, B. A.; Gregerman, S. R.; Jonides, J.; Von Hippel, W.; Lerner, J. S. Undergraduate student-faculty research partnerships affect student retention. Rev. Higher Educ. 1998, 22, 55–72. 36. Melendez, M. C.; Melendez, N. B. The influence of parental attachment on the college adjustment of white, black, and latina/hispanic women: a crosscultural investigation. J. Coll. Stud. Dev. 2010, 51, 419–435. 37. Mattanah, J. F.; Lopez, F. G; Govern, J. M. The contributions of parental attachment bonds to college student development and adjustment: a metaanalytic review. J. Couns. Psychol. 2011, 58, 565–596. 38. Collins, N. L.; Read, S. J. Adult Attachment, Working Models, and Relationship Quality in Dating Couples. J. Pers. Soc. Psychol. 1990, 58, 644–663.

28

Case Studies: Models that Improved Student Success

Chapter 3

A Comprehensive Model for Undergraduate Science Education Reform To Better Serve the Underserved Jim E. Swartz1,* and Leslie A. Gregg-Jolly2 1Chemistry

Department, Grinnell College, Grinnell, Iowa 50112, United States 2Biology Department, Grinnell College, Grinnell, Iowa 50112, United States *E-mail: [email protected]

Failure to succeed in introductory science classes is a barrier to diversification of the scientific workforce. In the early 1990s, it was found that Grinnell College students —particularly those of color, women, and first-generation college student—swere entering Grinnell College with an avowed interest in pursuing degrees in the sciences but abandoning their academic goals when they failed to do well in introductory science courses. To address this problem, a program called the Grinnell Science Project was developed to help students overcome three barriers to success in the sciences. We developed this list of barriers from data analysis of performance and issues experienced by students as: (1) unsuccessful acclimation to college life; (2) ineffectiveness of traditional pedagogy; and (3) a lack of mentoring and role models. Results of this project reveal improved grades for domestic students of color, as well as comparable rates of science major completion and pursuit of graduate study for all groups of students. The culture of the Science Division has changed to reflect, both in architecture and in actions, a commitment to establishing a supportive and inclusive community to promote excellent science. The sciences have undergone major curricular reform, including revision of introductory courses throughout the sciences to provide more active and engaged pedagogies and provide

© 2018 American Chemical Society

increased opportunities for course-embedded and dedicated research experiences.

Introduction Science is critical to modern society, for the role it plays in the economy, the provision, management and utilization of natural resources, and other benefits of technological innovation, especially those applied to healthcare (1). Scientific progress depends on having a well-qualified and ample workforce. According to reports from the President’s Council of Advisors on Science and Technology and the National Academy of sciences, the US is falling short of this demand. Since white men make up the majority of STEM (science, technology, engineering and mathematics) workers, a recommended approach to make up the shortfall is to make the sciences more inclusive in order to create a STEM workforce that closely matches the demographics of the population (2, 3). Women represent roughly 50% of the general population but only 25% of the overall STEM workforce (4). As described by leadership of the National Institutes of Health, certain racial and ethnic groups are severely under-represented in biomedical research. For example, scientific research faculty positions are only held by 4% African Americans, 4% Hispanics, 0.2% Native Americans, and 0.1% Hawaiian/Pacific Islanders (5) and little progress has been made to increase their collective proportion, despite increased representation in the US population (6). Reform efforts in undergraduate STEM education have been a critical target to improve recruitment and persistence of STEM workers (7). The fastest growing segments of our population, people of color, are least likely to access quality STEM education and most likely to leave STEM majors (2). An examination of the number of bachelor’s degrees awarded in STEM fields from 1977-2011 reveals that although there has been an overall increase in the proportion of women earning STEM degrees, this increase is largely accounted for by increases in the life sciences such that compared to men, the proportion of women earning degrees in the physical sciences and mathematics has slightly decreased since 2000 (8). Despite decades of attention, the low percentage of under-represented minorities among STEM graduates has persisted) (9, 10). A study conducted by the Higher Education Research Institute (UCLA) indicated that the primary cause of this under-representation is very low degree completion rates (11). Experiences in lower level science and mathematics courses are an important factor in STEM attrition (12). We need to target inclusivity efforts here to encourage and support students to ultimately complete STEM majors and pursue graduate study. Numerous national reports (13–15) offer a clear vision of a robust undergraduate STEM learning environment, including evidence that more engaged learning, such as active and inquiry-based learning, and research like experiences embedded in courses, are the kinds of activities that lead to better learning and retention in STEM (16–18). In response to data indicating the predominance of the disproportion of white men in the scientific workforce, Valantine and Collins (6) identify the need for scientific approaches to address challenges associated with diversity in science. 32

Among these recommendations, they call for evidence-based approaches to recruitment and training, including effective research experiences and providing mentoring; and interventions that mitigate individual and institutional barriers to effective education and workforce diversity (19). Handelsman (20) and Packard (21) have published work on developing effective mentoring to assist with this process. The work needed to best understand and evaluate ways to impact attraction and retention in the sciences will likely require examination and comparisons between several successful models across various institutions. Three recent publications, two focused primary on liberal arts institutions (22, 23) and another based on experiences at institutions in California, primarily comprehensive institutions from the California State University system (7), examine and describe processes related to transformational change efforts influencing whole programs or colleges. A number of inclusive models and case studies are available (AAC&U Diversity, Equity, & Inclusive Excellence Resource Hub at https://www.aacu.org/resources/diversity-equity-and-inclusive-excellence, for example), but few include analysis of the long term impact of these programs on student success, measured by outcomes such as proportions of STEM majors, post-graduate activities, and course grades. Researchers studying the outcomes of programs designed to improve persistence remain interested in the traditional analytic approach to parsing the individual contributions of treatment variables through both field studies and controlled experiments; however, a realistic understanding of the students’ experience may require a holistic approach to the program as a complex package of interacting treatments (24). Here, we describe a comprehensive set of interventions that have been developed and implemented in a coordinated fashion over more than a decade, and we demonstrate their effectiveness at improving rates of success for students from traditionally under-represented groups in the sciences at a small, highly selective liberal arts college. We call these cumulative efforts the Grinnell Science Project (GSP), and we examine the impact of the project by looking at measures associated with persistence in STEM fields. Some of the lessons learned from the GSP have been incorporated into a Web presence and publication, which are outcomes from institutions recognized by the Howard Hughes Medical Institute as “Capstone Awardees” (25) to recognize mature and successful programs making important contributions to undergraduate science education (22, 23).

Institutional Context Grinnell College is a highly selective, private, liberal arts college with about 1600 full-time undergraduate students. The College has a long history of commitment to equal access and to social justice. Grinnell practices a need-blind admissions process for domestic students and meets their full demonstrated need, resulting in over 90% of students receiving financial aid. Grinnell graduates over 40% of its students with science majors, and ranks seventh on a per-capita basis in producing graduates who pursue doctorates in the sciences (26). Grinnell asks students to take an unusually high level of responsibility for structuring their 33

education, while providing students with a mentoring relationship with a faculty advisor. Similar to the development of other programs designed to improve access to the sciences (7, 23, 27), our program originated with observations and an examination of data related to student preparation and success. In the late 1980s and early 1990s, a group of Grinnell College science faculty and student life professionals found themselves dissatisfied with the low numbers of domestic students of color graduating from Grinnell with majors in sciences. Data indicated that many such students with a declared interest in science completed majors in the sciences at far lower rates than did other students with science interests. We noted that average grades in introductory science courses were lower for students of color than for other domestic students. In fact, for African American students, the average grades in introductory level science and mathematics courses was a full GPA unit below majority students, and we looked, then, for a correlation between these students’ academic preparation and their grades in introductory math and science courses and found almost no correlation between these grades and their high school grades or standardized exam scores. We did find women earned better grades, on average, than did men, but women still did not persist in pursuing majors in the physical and computational sciences at the rate that men did. Further analysis of information from students’ applications for admission showed risk factors for poor grades to be: status as a first-generation college student, graduating from a high school where fewer than half the graduates enrolled in college, and being a domestic student of color. Thus, the factors interfering with academic success in the sciences were more likely to be social and environmental than academic. Our intervention strategy, then, needed to be more focused on social and environmental issues than on academic remediation. We also committed ourselves to a strategy of mutual adjustment, believing not just that the students needed to adapt to being at Grinnell College, but that we needed to change our approaches to more effectively assist students in excelling.

Objectives of the Grinnell Science Project (GSP) In addition to the national importance of increasing the size and diversity of the scientific workforce (2, 3), providing an environment in which sub-populations of students can achieve their educational goals is critical. Success in two elements of the Grinnell College mission—to appreciate and nourish diversity, and to provide an education that prepares our students to be contributing citizens in the world community depended upon improving our performance. We created the GSP to address the barriers that we found that inhibited the persistence of some students in STEM at Grinnell College and that, if successful, would more effectively meet the institutional mission. At the highest level, the GSP aims to enable all students’ full participation in the Grinnell community, including pursuit of science majors and careers. We targeted our efforts at students who come to Grinnell planning to pursue science and mathematics but who are among the traditionally underrepresented populations in science and mathematics, or who are identified by our data analysis as being in a group “at risk.” Since we had a long tradition of success 34

in graduating students in STEM, just not as diverse a group as desired, we felt well poised to take on this project to spread that excellence to a group that better represented our student body and the national needs. Based upon our data analysis and the work of Treisman (28), we identified three barriers to the pursuit of a science or mathematics major and to full participation in the community. These are: acclimation to college life, ineffectiveness of traditional pedagogy, and lack of role models and contexts for the study of science. The GSP includes program objectives designed to respond to each barrier: • • •

To foster acclimation to college life through providing a pre-orientation; To improve curricula and pedagogy through creating interactive science/ mathematics courses and increased opportunities for mentored research; To provide role models and contexts for the study of science through mentoring and community building.

These elements are consistent with Graham’s persistence model published subsequently (16). More recently the community has begun to pay attention to social science work that appears to offer a better understanding of factors leading to the lower rate of persistence in STEM by URM students including, stereotype threat (29, 30) growth mindset (31, 32) and micro-aggression (33–35) which play a significant role in decreased persistence in STEM, particularly at the undergraduate level. Meeting the objectives of GSP required comprehensive changes. As described below, program elements were created in response to our local data, other efforts for improved pedagogies and inclusive practices available at the time, and outside speakers and consultants. Although a theoretical model for the totality of the changes was not followed, the approach and activities of GSP align closely with recent recommendations (36) for increased diversity and persistence in STEM based on Lewin’s theory of “planned approach to change” (37, 38). These recommendations were made by a Joint Working Group on Improving Underrepresented Minorities (URMs) Persistence in Science, Technology, Engineering, and Mathematics (STEM) convened by the National Institutes of General Medical Sciences and the Howard Hughes Medical Institute. At their core, the recommendations support Lewin’s point that change requires changes at the large system or group level, as opposed to shifting individual behavior. According to Lewin, objective standards of achievement are necessary to facilitate a sense of achievement and learning (37). Revisiting hallmarks of achievement are motivational and allow for a spiral approach where evaluation influences planning, and execution. Revisiting data after execution then can inform planning and execution, and so on. This collection and utilization of relevant data is critical for “unfreezing” the status quo and promoting change. Therefore the first and essential step of the Joint Committee recommendations includes the evaluation of the institution regarding equity and inclusion by collecting and evaluating data tracking degree candidates and earners in STEM disciplines across demographic categories (36). As was the case for the origin of GSP, such data were essential for publicizing and grappling with 35

the discrepancies in achievement between demographic groups and provided information (“diagnosis” in Lewin’s terms) for shaping the programmatic elements to address the inequities (“treatment” in Lewin’s terms). The data were critical in the design of the GSP program elements, directed by intervention strategies on social issues. Periodic review of data impacted programmatic elements, such as eventually dropping women in chemistry as a targeted group for inclusion in the pre-orientation since success of that group rose to be on par with all students (described below).

Program Elements To Enhance the Participation of Members of Groups Under-Represented in STEM In order to plan and implement the program, we engaged both faculty and student affairs professionals and relied upon advice from students who were members of our target group. A small group of faculty members, student affairs leaders, and a grant writer in the Dean’s Office provided leadership by engaging the broader group of faculty and staff in planning and carrying out efforts and in providing faculty development activities. We invited individuals who had undertaken impressive efforts elsewhere as speakers and consultants, and drafted grant proposals both to provide seed funding and to help focus our objectives and efforts. Key grants (see Grant Support section, 1992-94) provided support for the pre-orientation program, hiring term faculty members so that we could offer smaller experimental sections of introductory courses, and engage students from the target group in research projects. Receipt of the grants not only provided financial support, but also kept us focused upon accomplishing what we set out to do. We also put in place a system of gradually phasing out grant support and phasing in college funds for ongoing expenses, so that financial support for the activities did not end at the end of the grant funding. These strategies are parallel to those subsequently recommended by Henderson and colleagues in 2011 (39).

Component 1: Pre-Orientation Objective: To Foster Acclimation to College Life. We designed a one-week pre-orientation, intended to build confidence and to alleviate the anxieties of the first year, since these may create an uncomfortable campus climate for a student and hinder his or her academic performance. This pre-orientation is held the week preceding general orientation for new students. Using the students’ college applications and their transcripts, students with risk factors (from our data analysis described above) for poor performance in introductory science courses (none of which involves academic preparation) are identified. The selection of these targeted students is based on their being one or more of the following: •

a first-generation college student; 36

• •



a graduate of a high school where fewer than half the graduates enrolled in college; a domestic student of color (Hispanic, African American, Native American/Pacific Islander, or Asian American, who were underrepresented in STEM at Grinnell); and/or a woman interested in physical and computational sciences.

From that group, students who have indicated an interest in science on their applications and have accepted admission to the College are invited to participate in the pre-orientation. This has resulted in participation of 25-35 students in the early years, and 40-50 in recent years. These selection criteria have changed modestly over time. For instance women in chemistry were originally included, but were then omitted all as the number of graduates reached parity with men, and originally including all Asian Americans, but recently only those who otherwise qualify. The aims of the pre-orientation include: • • • • •

providing a small student cohort in which relationships and a support network may be built; creating supporting relationships between students and a variety of science faculty; acquainting students with an array of support services the College provides; identifying particular academic or writing weaknesses; and helping students become comfortable with the campus geography, the library, the computers, and residential life.

According to a national survey of four-year colleges and universities across the U.S. by Barefoot, Griffin, and Koch (40), most (93%) bridge programs strive to support student development of academic readiness. This element is not part of the GSP pre-orientation due to the focus of our intervention strategy on social issues. However, strong faculty involvement and close and frequent interactions with faculty from across the Science Division is a key feature, one which makes it distinct from our general first-year orientation. These students meet faculty members who teach introductory science and mathematics courses and hear faculty expectations for students in these courses. Students also participate in facultyled sample classes and a research-like project. Additionally, faculty members participate in many of the social events, starting by dining with the students and the families when they arrive on campus. Six to eight student assistants help with both the academic and the residential-life aspects of the pre-orientation. Since Grinnell asks students to take an unusually high level of responsibility for their college experience, both in the academic and student life realms, developing a comfort with using faculty, staff and more experienced peers as mentors is crucial to success. Developing those relationships is a key goal of the pre-orientation. The GSP students and their families, many of whom are involved in college education for the first time, have the opportunity to meet other students, learn about the services and structures of the college, and meet faculty. The 37

target population consists of students who express an interest in science and mathematics, indicating that they come with common interests. The smaller size of the group, compared to the general orientation, fosters student-student and faculty-student relationships that help build a supportive mentoring network for the next four years. The burdens of adjusting to the many new demands made on the students are relieved by the personal attention they receive during the pre-orientation. Furthermore, the participants are local “experts” when other new students arrive, and can offer directions and advice to them, further bolstering their confidence. Thus far, over 900 students have participated in the pre-orientation program. Over time, the pre-orientation successes have informed and helped us to improve our general student orientation programs, which are primarily run by our student life offices. Additional information about the pre-orientation program can be found at https://serc.carleton.edu/lsamp/bridging/programs/grinnell.html. Component 2: Engaged Learning in Science and Mathematics Courses Objective: To Improve Pedagogy, Especially in Introductory Science and Mathematics Engaged pedagogies (41) are a key element for efforts at inclusion in the sciences (16, 42). A leader in mathematics education for under-represented students, Professor Uri Treisman, named three main elements of interactive learning: (1) a focus on helping participants excel, rather than merely to avoid failure; (2) an emphasis on collaborative learning and the use of small-group teaching methods; and (3) faculty sponsorship, needed to nourish the program and enable it to survive (28). Traditional classroom methods have proven to hinder academic success for the target population (43). When faculty use active learning strategies in the classroom, as opposed to traditional lecture style of content delivery, student failure rates decrease and exam scores increase (44). Our curricular and pedagogical changes remove the burden of adapting to traditional styles of learning. This work has evolved over the past 25 years, and the changes have proved far more pervasive, as described below, than we originally imagined. The curricular and pedagogical development component of the program has aimed at changing the basic fabric of introductory courses by providing faculty members with a nurturing environment, mentoring, and the intensive development time they need to make such changes. The President and the Dean of the College supported this work, both intellectually and financially as well as in the faculty reward system, with a special emphasis on supporting the integration of teaching and research. The goals of our curricular and pedagogical changes are to provide challenging, not remedial, problems which engage the students in hands-on investigation and mutual realization of the solutions, and to respond to different learning styles. We started with several experiments. We launched a series of one-credit (the regular course is four credits), add-on courses that students could co-enroll in with the standard introductory courses. These one-credit courses provided students with interactive ways of learning material, and provided platforms both for more engaged learning by students who wanted some 38

additional work, and for pedagogical experimentation by faculty members. The experimentation by faculty members allowed one or more individuals to try some engaged learning and investigative approaches without having to convince all course instructors or to give up class time in the regular course. In addition, our physicists decided to experiment with a variation of the workshop (no lecture) physics approach pioneered by Priscilla Laws (45) for one of the three sections of the introductory physics sequence, and the entire computer science faculty decided to transform their introductory courses into a workshop format. After two to three years, faculty members became convinced that learning improved dramatically as a result of these experiments (46). As a result, the engaged pedagogies were integrated into the standard courses, and the one-credit courses were abolished. In physics, roughly half the students now opt for the workshop format and half for the more standard lecture-lab approach. The first course in computer science is taught entirely in the workshop format, and substantial portions of many other CS courses use active learning approaches. The introductory biology course is entirely based on a research project (47, 48). Some chemistry sections are taught entirely in a workshop format and others use many engaged learning techniques, including research-like projects and learning in the context of a social problem (global warming, water quality, etc.). Psychology and mathematics also use a number of engaged learning approaches. In all, these changes, along with increased levels of student-faculty research, substantially increased faculty mentoring of students at all levels of the curriculum. In addition to new pedagogical approaches, the sequence and content organization of the 100- and 200-level biology and chemistry courses have been revised to ease the scheduling and performance pressures on first year students. Resulting from a decade of emphasizing curricular reform and active learning in the science classroom, departments across Grinnell’s Science Division teach groundbreaking introductory courses. Where many biology departments are struggling to fit more and more material into the introductory course, and only a few have even broached the idea of workshop-style teaching, Grinnell’s Biology Department has decided that the most important learning outcome of the introductory course is to get students to think like a biologist. Students in the introductory course read original research papers, design and conduct their own experiments, analyze data, and present results in forms appropriate to the discipline, including posters and research papers. The resulting course aligns very well with recommendations issued in the 2011 report Vision and Change in Undergraduate Biology Education: A Call to Action (49). Where many computer science departments continue to disallow students working together, thereby discouraging students who value teamwork, Grinnell’s Computer Science Department embraces collaboration in a wide variety of courses, particularly in the introductory course, which relies almost exclusively on workshop-style exercises throughout the course. By participating in science education that is structured much more like how science is practiced, students are engaged in the practice of science and the relationship with the instructor becomes a mentor-apprentice relationship. New courses and major course revisions are shown in Table 1. While there were few research-based introductory science courses in the early days of the GSP, they have become more common and a 39

recent publication provides substantial evidence for their efficacy in improving degree completion in STEM (50).

Table 1. Major Course Revisions Linked to GSP Objectives Course

Title

Revisions

Year Implemented

Introduction to Biological Inquiry

NEW, completely inquiry/research-based, workshop format

Fall 2001

Biology 150

Biology 251

Molecules, Cells, and Organisms

NEW, content rearranged, small group work

Fall 2001

Organisms, Evolution, and Ecology

NEW, content rearranged, workshop format

Spring 2002

Biology 252

Biological Chemistry 262

Introduction to Biological Chemistry

NEW, introductory course taught jointly by biology and chemistry faculty

Spring 2001

General Chemistry

Content rearranged, hands-on modules, small group work, workshop section

Fall 1997

Organic Chemistry

Engaged pedagogies, small group work, workshop section

Fall 1999, Fall 2008Workshop

Fundamentals of Computer Science

Workshop style, transition to multi-paradigm approach instead of the more traditional imperative paradigm approach

Fall 1992

Functions and Differential Calculus, Functions and Integral Calculus

NEW, slower paced equivalent to core calculus semester

Fall 1997 / Spring 1998

General Physics

workshop sections

Spring 1993, Fall 1994

Psychology 113

Introduction to Psychology

NEW, smaller sections, more experimentally based work

Fall 2002

Mentored Advanced Projects

NEW, integrate student-faculty research collaborations into curriculum

2002

MAPs 499

Chemistry 129

Chemistry 221-222 Computer Science 151-152

Mathematics 123-124

Physics 131-132

Various formal and informal gatherings of science faculty reflect upon our curriculum and pedagogy. Regular meetings of a Science Teaching and Learning Group attract a wide range of faculty, from new faculty to full professors with many years of experience. These provide a venue for mutual mentoring among 40

more experienced faculty members and those new to Grinnell College, among those who have tried innovations and those just getting started, and among newer faculty members emerging from a different pedagogy and those long committed to a didactic style.

Component 3: Establishing a Community Objective: To Enhance Mentoring, Collaborative Work, and Students’ Self-Identification as Scientists Science and the study of science are facilitated by communities of practice. One of the classic barriers for first-generation college students and students of color is feeling they do not belong in those communities (51). Building community and a sense of belonging is a key goal of the pre-orientation program, but our attempts to build community are not limited to that. They include much group work in the context of courses, in student-faculty research, and continued support for mentoring relations that give students genuine experiences teaching them to act and feel like they are scientists working within a scientific community. All these and other efforts increase students’ confidence, preparedness, achievement, and ownership of the community.

Student-Faculty Research A critical element of mentoring and feeling like a scientist is student-faculty research. There is a growing body of research documenting how students benefit from research) (52–54). Faculty-student research increases persistence of students (55, 56), and early research experiences seem to be especially effective (57). We recognized, as have many others, that student-faculty research is one of the best forms of mentoring. We have incorporated small research projects and research-like experiences into many of our courses, including first and second year courses, so that students experience science as it is done and build community with their fellow students as they work together on such projects. We have increased the number of students in our academic year and summer research programs, and regularized the selection process, including a bias in favor of members of the at-risk group. When possible, we have sought to involve members of our target group in research projects early in their undergraduate careers (56). Based upon our own experience and on what we have learned from our ROLE project led by David Lopatto (53), we have helped faculty understand ways they can best support student learning in research. We have instituted a formal process of faculty evaluation of students and student evaluation of the program, to assist us in improving effective mentoring. In the late 1990s, in part based upon our successes in the sciences, we established a college-wide program, not limited to the sciences, named Mentored Advanced Projects (MAPs). We also have established department- and division-level activities, such as informal lunches and gatherings with faculty members, focused upon helping students to 41

see themselves as scientists and to help create community. In a typical year, close to 200 student MAP projects are completed in the sciences. Faculty evaluations report that fewer than 10% of these students need close supervision and regular direction, and that over 90% of the students could relate their findings to the context of the scientific work. In a separate project, Jerod Weinman, David Jensen, and David Lopatto found that students’ self-reported learning gains from research experiences in computer science improved when these experiences included not only instruction in the important mechanics of research but also grounding in a philosophy of computing science that emphasizes generalized explanation of behavior as a means for control and prediction (58).

Establishing a New Science Learning Center (SLC) and Math Lab There is substantial evidence that mentoring is a key component to success and retention of students, particularly those who are members of groups under-represented, in STEM (21). In 1997 we created the SLC to provide one-on-one tutoring in the introductory science classes and supervision of peer mentors in science courses. (The College already had a long-established Math Lab and a peer mentoring system in biology and chemistry (59).) It was clear to us that our peer mentoring and tutoring efforts could be improved with some professional leadership. It was also clear that we had something of a gulf between the faculty and students with respect to communication about what was effective and ineffective in our classrooms, and how improvements might be made. Students knew that they were having problems, but they did not have the experience to tell us what to do about it. Hiring a professional director for the SLC, who has both a background in science and in tutoring and mentoring, as well as an awareness of pertinent literature, of interesting projects elsewhere, and of the institutional barriers to reform, has been a tremendous boost to our reform efforts as well as to tutoring and campus climate issues. She has had a strong impact on our project, not only by working effectively with both students and faculty, but also by helping us understand what is needed institutionally to support reform. The SLC provides one-on-one tutoring for students, training of the peer mentors, and working with faculty, individually and collectively, to improve the learning environment for students. Most first- and second-year courses employ an experienced student peer mentor. Mentors are selected based on their own performance in the course, and on an application that indicates their proposed approach to mentoring, and what they, personally, bring to the goals of the mentoring program. The mentor sits in on all class sessions, assists with active learning experiences, and runs support sessions outside of regular class meetings. The SLC provides pedagogical training for the mentors and a structure for the mentor to provide weekly feedback to the instructor in the course. Faculty have noted that this relationship is really one of co-mentoring, involving the SLC director, instructor, and peer mentor, all aiming to improve student learning. Many of the mentors have been from our target population of students at risk; thus, besides providing additional professional development for the mentors 42

themselves, they can model success to students enrolled in the courses. Mentors report that this experience enhances their learning and aids them in seeing themselves as scientists (60).

Building Community: Space as a Critical Factor of Creating Communities The Robert N. Noyce ’49 Science Center was completely renovated and expanded in two phases, the first completed in 1997 and the second in 2008. The Center houses all of the science departments—biology, chemistry, computer science, mathematics and statistics, physics, and psychology—as well as the Kistle Science Library and dedicated spaces for the Science Learning Center and Math Lab. Research laboratories have been updated with excellent facilities and equipment. They are also easily linked with classrooms. Based upon our curricular changes, teaching laboratories are designed more like research laboratories, often optimizing cooperative learning. With the increased numbers of workshop-based courses, lab spaces make it easy for students to move between class and lab. Interior glass brings natural light into the classrooms and labs, and increases visibility of teaching and research to those walking by. There are spaces for various sized groups of students to work and study. As reported in Gregg-Jolly et al. (61), focus groups investigating both challenges and supports for STEM students, they noted (without prompting) that the design of the Noyce Science Center helped to create a supportive community: “I think Noyce is extremely excellent for supporting collaboration and openness, because there’s just like tables, and you just like walk by, and you see someone at a table, and you chat about the homework. So I think the communal spaces here really do a lot to help that.” “I can just go to the computer science lounge and see other people and chill with them. Having more spaces where people can have that interaction – that tends to make people feel like they belong.”

The Second Year More recently, we noted that first-generation college students and domestic students of color at Grinnell exhibited a lower level of academic success in second year science courses than other students did. We conducted a mixed method assessment of this work involving both a survey as well as focus groups and interviews of second year students enrolled in second year science courses. We implemented a number of programs to try to improve our success, including a second year retreat for second year science students, faculty development workshops on group or cooperative learning, meta-cognition, and implicit bias. After four years, the success rate of domestic students of color and first generation students improved to nearly that of other students. This work is described in a recent publication (61). 43

A Web of Mentoring Building on Grinnell College’s program of close student-faculty interactions in our individually advised curriculum (without core or distribution course requirements), we have created a highly interconnected web of mentoring (Figure 1) that supports new students in many directions so that success is not dependent upon a single element. This web also exemplifies our approach of changing the institution to support student success, instead of simply expecting students to adapt to the institution. For example: • • • • • • •

• • • •

Pre-orientation establishes faculty mentors for new students; Pre-orientation provides experienced and accessible student mentors for new students; Pre-orientation establishes supportive relationships among new students; Follow-up social and informational events with the pre-orientation group further build community and confidence ; Interactive classes promote further faculty mentoring for new students Cooperative learning strengthens mutual peer support; Use of mentors in courses provides experienced student support for new students and helps the new students feel like they belong in the community of scientists at Grinnell College; Training of mentors establishes SLC director as support for experienced students and the mentors in co-mentoring relationships with one another; Course mentoring establishes co-mentoring relationships between course instructors and course mentors; Research projects at multiple levels of the curriculum provide faculty mentoring of research students; Science Teaching and Learning Group and other activities promote comentoring among faculty.

Outcomes Over the past 20 years, the Grinnell Science Project has been remarkably successful on multiple fronts. Over 900 students have participated in the pre-orientation program and thousands have benefitted from the curricular and pedagogical changes as well as mentoring relationships that have been established by the GSP. Another measure of success for the GSP program is that, whereas for women, an interest in chemistry was originally a selection criterion for invitation (due to substantial underrepresentation), the number of women chemistry majors is similar to or exceeds the number of men so that it is no longer necessary to target them specifically. We used official IPEDS race and ethnicity categories as self-reported by our students for our demographic data. There was a change in those official categories during the time of our data reporting, and we have done the best that we can to map them to one another, but the largest impact was creation of a new category of two or more. 44

Figure 1. Web of mentoring. Illustration of the comprehensive support network provided for the GSP student. Effective collaboration and coordination between interconnecting elements strengthens the network. Even if a particular element fails, the student is still supported.

Grades In our first data analysis, we found that African American students’ grades in introductory science and mathematics courses lagged behind other students by one full grade point. This was similar to findings by Treisman in calculus courses at the University of California at Berkeley (28). Other domestic students of color lagged significantly, but by somewhat lesser amounts. During the first few years of the GSP, we found that the grades of students who participated in the pre-orientation program were substantially higher than those who were invited but did not participate. Even though the results were very positive and indicated success, we abandoned this comparison after several years, for two 45

reasons. First, we surmised that there was likely a sample bias between the students who elected not to participate in the pre-orientation and those who did, with those electing non-participation being less committed to the study of science. Second, as the pedagogical and curricular changes became more pervasive, it was not clear that participation in pre-orientation was the main variable accounting for any differences. We now compare the average grades of domestic students students of color (the highest risk factor for poor grades) to those of all students in introductory math and science courses (those that are gateways to STEM majors). We have done this comparison for two multi-year periods in the past decade, and we find that difference in the averages grades of the two groups has shrunk dramatically from roughly 1.0 GPA unit, to about 0.2 GPA units.

Science Graduates In addition to improving the academic performance of students, so that they could pursue their educational and career aspirations, we hoped that this improved academic performance (grades) along with more exposure to science and to scientists as role models, would encourage them to pursue science majors. The results are impressive. •

• •





Prior to the GSP, from 1992–1994, the College graduated an average of 42 science majors annually who were women and eight who were students of color. By 2015, those numbers had jumped to 77 women (an 82% increase) and 33 students of color (an over 300% increase). Three African American women majoring in physics were among the Grinnell College class of 2008. Comparing students (who started in 1998–2000) who participated in the pre-orientation to those who were invited but declined, 79% of the GSP participants eventually declared a major in science, while of those who did not participate, only 39% did so. For women, the proportion is 61.5% to 37.9%; for African Americans it is 60.5% to 27.8%. We note that these data may be affected by the selection process for inviting participants and their decisions to participate or not. The proportion of physical and computational science majors who are women nearly doubled from percentages in the low 20s in 1990-94 to over 40% currently. The percentage of science graduates who are first generation college students increased from 12% in 2004-08 (the first years we have comparable data) to 16% in 2014-16.

The changes in the number of science graduates are shown in Figures 2-4. As can be readily seen the number of science majors graduating from Grinnell College has nearly doubled over the past 25 years, but the increase has been nearly all non-white students, with the proportion of white graduates decreasing from 80% to less than 60%. 46

Figure 2. Number of science graduates at Grinnell College for class years 1991-2015. Data for all science majors (biology, biological chemistry, chemistry, computer science, general science, mathematics, physics, and psychology) graduating with a BA degree in the academic year ending in the years indicated. The majors are aggregated by self-reported race and ethnicity: White, Foreign National (FN), African American (Af Am), Hispanic (Hisp Am), Asian American (As Am), Native American (Nat Am), Unknown (Unk), and two or more, 2+. (see color insert) As reported by the National Science Foundation (26), Grinnell ranks seventh on a per-capita basis among all U.S. higher education institutions in producing science graduates who go on to pursue a Ph.D. Recent graduates who participated in our GSP pre-orientation program were more than twice as likely to pursue graduate study in STEM fields compared to those who did not participate. The GSP is a first step for many of these future scientists. Curricular and Pedagogical Change One of the most enduring changes that the GSP has nurtured is a Science Division wide commitment to and faculty development opportunities focused upon curricular and pedagogical reform (Table 1). Through building a community of practice that supports engaging and inquiry-based pedagogical approaches, the students experience a curriculum that is focused upon providing excellent learning experiences for them. Student-focused reflection and continuous improvement are the norm, as evidenced by the fact that nearly every faculty member in all six science departments has implemented some change as a part of this community. 47

There are roughly monthly lunch meetings called the Science Teaching and Learning Group. Nearly every summer there is a focused workshop on topics like cooperative learning, metacognition, and stereotype threat. We believe that this supportive community of practice with a strong commitment to mentoring of students has enhanced our ability to attract and retain a strong group of new faculty as well. In 2017, six out of 54 tenured and tenure-track science faculty members are domestic faculty of color, whereas in the mid-1990s none were. Becoming a Professional In a review of the Science Learning Center and the mentoring program, a survey of alumni who were mentors provided some surprising results (60). More than 75% of the mentors reported the following gains from serving as a mentor: exposure to multiple pedagogies, appreciation for active learning, increased confidence in the subject material and presentation skills, and an increased sense of responsibility and of being a role model. These gains clearly have an impact on the future careers of the mentors: 87% of the respondents were working in science, teaching, practicing medicine, or in graduate training to do so.

Figure 3. Number of science graduates of color at Grinnell College for class years 1991-2015. Data for science majors (biology, biological chemistry, chemistry, computer science, general science, mathematics, physics, and psychology) graduating with a BA degree in the academic year ending in the years indicated. The majors are aggregated by self-reported race and ethnicity: African American (Af Am), Hispanic (Hisp Am), Asian American (As Am), Native American (Nat Am), and two or more, 2+. (see color insert) 48

Figure 4. Percentage of ethnic groups making up science graduates at Grinnell College for class years 1991-2015. Data for all science majors (biology, biological chemistry, chemistry, computer science, general science, mathematics, physics, and psychology) graduating with a BA degree in the academic year ending in the years indicated. The majors are aggregated by self-reported race and ethnicity and converted to the percentage of the graduates in that period that fit into each category: White, Foreign National (FN), African American (Af Am), Hispanic (Hisp Am), Asian American (As Am), Native American (Nat Am), Unknown (Unk), and two or more, 2+. (see color insert) In a review of all (over 2400) science major graduates from 2000 through 2014 (see Table 2), we found that 54% of graduates had earned or were pursuing a STEM graduate degree. That percentage for domestic students of color was also very high, 50%. Students of color were a bit less likely to be pursuing or to have completed a STEM PhD and a bit more likely to be pursuing or to have completed a health-related graduate degree. We do note that students of color are substantially more likely to be first generation college students, and that might influence their choice of graduate study in STEM or health care. Sustainability The aspects of the GSP that have proven successful have been incorporated into the College’s standard operations and budget. The costs represented in the budget for at least the past 15 years include an expanded tenure-track faculty, a full time SLC director, student staff (mentors and pre-orientation student staff), faculty development activities, and the direct costs of the pre-orientation and curricular changes. Certainly, our experience and successes stimulate new ideas, and these are cited in grant proposals to support new curricular and pedagogical projects. 49

Table 2. Outcomes of STEM Graduates (2000-2014) from Grinnell Collegea Student Characteristic

Any STEM Graduate Study

STEM Ph.D.

Health Graduate Program

All STEM graduates

54%

24%

12%

SOC

50%

18%

15%

FN

56%

24%

9%

White

55%

25%

13%

Other

50%

24%

12%

First generation

43%

16%

12%

Non-first generation

55%

24%

13%

GSP participant

51%

22%

12%

a

The graduates are for all science majors (biology, biological chemistry, chemistry, computer science, general science, mathematics, physics, and psychology) graduating with a BA degree in 2000-2014. The majors are aggregated by self-reported race and ethnicity(White, African American (Af Am), Hispanic (Hisp Am), Asian American (As Am), Native American (Nat Am), Unknown (Unk), and two or more, 2+) Foreign National (FN), or first generation.

More critical and impressive is that the GSP is simply part of the Grinnell College faculty and administrative culture. Leadership is critical to reform efforts (7), but that does not mean that the same person or group of people have to continually lead. In the GSP, we found that changes in faculty and student-life leadership have strengthened the program, rather than weakened it. Faculty directors have included members of all of the six science departments. Well over half of the science faculty members, representing all six departments, have participated in some aspect of the pre-orientation program. It is common for a faculty member to serve as a resource person for one or two years, lead one of the research projects the following year, then serve as an assistant director and eventually a director. Thus, we have built-in faculty mentoring and leadership development, with a constant renewal of engaged faculty members. Faculty experience with this program supports a culture of pedagogical reflection to support curricular and pedagogical change.

Grant Support External grant support (Table 3) has done a number of critical things for the project. It forced us to plan and articulate clearly the needs of the project. It provided critical peer review and external validation of our efforts. Finally, it provided seed and developmental funding for GSP activities. These activities are now built into the College’s base budget and sustained by internal College funds.

50

Table 3. External Support for GSP and Related Activitiesa Year 1992

Organization GTE Foundation

Amount $30,000

PI

Title Minority Undergraduates in SMET

Swartz

1992

The Lilly Endowment

$150,000

The New Science

1993

GTE Foundation

$30,000

Factor

Swartz

1994

National Science Foundation

$148,683

Multidisciplinary Interactive Introductory Science and Mathematics Reform at Grinnell College

Schneider

1994

National Science Foundation

$49,993

“Learning Chemistry by Doing What Chemists Do”

Swartz (co-PI)

1995

GTE Foundation

$30,000

Focus “Peer Scientists”

Swartz

1995

National Science Foundation

$20,616

A Computational Physics Laboratory and Course

Schneider

1995

National Science Foundation

$2,871,500

ChemLinks: “Making Chemical Connections” (comprehensive reform of chemistry curricula; consortium grant)

Swartz (co-PI)

1996

National Science Foundation

$196,883

Institutional Reform

Duke

1998

National Science Foundation

$500,000

Award Integration of Research and Education

Swartz

1998

National Science Foundation

$75,000

CCLI/Introductory Biology

Robertson

2000

Howard Hughes Medical Institute

$900,000

Undergraduate Science

Voyles

2000

National Science Foundation

$651,885

ROLE/Assessment of Student Research

Lopatto

2004

Howard Hughes Medical Institute

$1,400,000

Undergraduate Science

Lindgren

2003-08 Howard Hughes Medical Institute

$445,029

Assessment of Student Research

Lopatto

2008

Howard Hughes Medical Institute

$1,200,000

Undergraduate Science

Levandoski

2011

National Science Foundation

$10,000

Presidential Award for Excellence in Science, Mathematics and Engineering Mentoring (for the Grinnell Science Project)

Swartz

2012

Howard Hughes Medical Institute

$1,000,000

Undergraduate Science

Gregg-Jolly

a

Projectb

Schneider

Excludes grants with research or equipment acquisition as primary objective. Grinnell Science Project (GSP).

51

b

Now

Conclusion By promoting student inclusion, achievement and excellence in teaching and learning, the Grinnell Science Project addresses the issue of increasing the diversity of the STEM workforce. The project’s multifaceted approach uses a range of activities rooted in intensive mentoring and building a community of scientists (students and faculty alike) along with associated curricular reform that supports persistence in science through and after graduation (16). Over 900 students have participated in the pre-orientation program and thousands of other students have benefited from the curricular and pedagogical changes, as well as from the various activities that we have developed to help build a supportive community. Prior to the GSP, from 1992–1994, the College graduated an average of 42 science majors annually who were women and eight who were students of color. By 2015, those numbers had jumped to 77 women (an 82% increase) and, as shown in Figure 2, 33 students of color (an over 300% increase). Since the inception of the GSP, grades for domestic students of color in introductory STEM courses have improved markedly to nearly on a par with other students. GSP alums have established a tradition of high achievement resulting in successful careers as research scientists, mathematics and science educators, physicians, and a few are contributing in other ways, for example, working as patent attorneys or, in one case, as Associate Dean of Admissions & Director of Multicultural Admission. The community of faculty and staff resulting from GSP-related efforts is attentive to practicing what we now call engaged and inclusive pedagogy and is committed to an engaging science and mathematics curriculum rich with collaborative research experiences. We believe that two hallmarks of success of the program were the recognition that 1) we needed to alter what Grinnell College did to take advantage of the distinctive strengths that our students brought to the College and 2) to help build a supportive and engaged community, and that the program is comprehensive dealing with both social and academic aspects of the college experience. The development of mentoring strategies and the curricular pedagogical changes that occurred were stimulated by pioneering changes by faculty members at other institutions (often unpublished), using approaches that are now, more than 20 years later, well documented as effective strategies for improving student learning and retention in STEM (16, 23, 36, 41, 42). All Grinnell science students have benefitted from this community, and this approach to curricular development and mentoring has especially benefitted the groups of under-represented students who participate in the GSP pre-orientation program. Another critical factor in our success was the repeated use of data. First, data revealed the problem was a social/environmental one, rather than the presumed lack of academic preparation of particular students. Then, data informed a change in our selection criteria for the pre-orientation program as the success rates of our students as well as the demographics of our student body changed. The use of data also helped to bring along faculty members and others who might have initially been pessimistic about whether we had a problem, the nature of the problem, and that changes could result in improved outcomes. The GSP provides a successful model for engaging students and faculty in a meaningful way and for attracting and training future scientists that is ripe 52

for adoption by other institutions. Perhaps more importantly we provide data regarding long-term impact on persistence that supports the efficacy of those efforts along the lines called for by Valantine and Collins (6). In 2011, the College received the Presidential Award for Science, Mathematics and Engineering Mentoring. We have worked with representatives of several other national liberal arts colleges (22, 23, 62), as well as a regional NSF LSAMP group (63) to collect and disseminate successful strategies to encourage and enable a more diverse array of science students at our colleges.

Acknowledgments Although the two authors of this paper did prepare the manuscript, the work described here represents the efforts of roughly 100 faculty and staff members at Grinnell College over the past 25 years. In particular, Mark Schneider, Professor of Physics, Anita Solow, Professor of Mathematics, and Jo Calhoun, Dean for Academic Advising, were key early leaders of the project. Joyce Stern, Dean for Student Success and Academic Advising, Clark Lindgren, Professor of Biology, Elaine Marzluff, Professor of Chemistry, and Minna Mahlab, Director of the Science Learning Center have been critical leaders over the past 20 years. We are grateful to their commitment as well as for that of our students, who have been articulate in expressing their needs and complimentary on where we succeed and helpful on where we could improve. The College administration was helpful in providing consistent support, both in terms of words of encouragement and appreciation, as well as substantial financial commitments. We thank Susan Ferrari for providing statistics on outcomes of recent STEM graduates, Carlie VanWilligen for assistance with data acquisition, and Erika Jack for essential contributions to preparing this manuscript. Grant funding, as outlined in Table 3 has also been critical in support of the project.

References 1. 2.

3.

4.

Xie, Y.; Killewald, A. Is American Science in Decline? Harvard University Press: Cambridge, MA, 2012. President’s Council of Advisors on Science and Technology. Engage to Excel: Producing One Million Additional College Graduates with Degrees in Science, Technology, Engineering, and Mathematics. http://www.whitehouse.gov/sites/default/files/microsites/ostp/pcast-engageto-excel-v11 (accessed December 11, 2015). National Academy of Sciences, National Academy of Engineering, and Institute of Medicine. Expanding Underrepresented Minority Participation: America’s Science and Technology Talent at the Crossroads; National Academies Press: Washington, DC, 2011. Beede, D.; Julian, T.; Langdon, D.; McKittrick, G.; Khan, B.; Doms, M. Women in STEM: A gender gap to innovation. Econ. Stat. Admin. 2011, 4–11. 53

5.

6. 7.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

National Science Foundation, NCSES. Women, Minorities, and Persons with Disabilities in Science and Engineering; Special Report NSF 15-311; Arlington, VA, 2015; Table 9-30. Valantine, H. A.; Collins, F. S. National Institutes of Health addresses the science of diversity. Proc. Natl. Acad. Sci. U.S.A. 2015, 122, 12240–12242. Elrod, S.; Kezar, A. Increasing Student Success in STEM: A Guide to Systemic Institutional Change; Association of American Colleges & Universities: Washington, DC, 2016. Mann, A.; DiPrete, T. A. Trends in gender segregation in the choice of science and engineering majors. Soc. Sci. Res. 2013, 42, 1519–1541. National Research Council. Expanding Underrepresented Minority Participation: America’s Science and Technology Talent at the Crossroads; The National Academic Press: Washington, DC, 2011. National Science Foundation. Women, Minorities, and Persons with Disabilities in Science and Engineering; NSF 17-310. https://www.nsf.gov/ statistics/2017/nsf17310/ (accessed January 2017). Higher Education Research Institute at UCLA. Degrees of Success: Bachelor’s Degree Completion Rates Among Initial STEM Majors, 2010. https://heri.ucla.edu/nih/downloads/2010-Degrees-of-Success.pdf (accessed January 22, 2018). Chen, X. STEM Attrition: College Students’ Paths Into and Out of STEM Fields; NCES 2014-001; National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education: Washington, DC, 2013. Fox, M. A.; Hackerman, N. Evaluating and improving undergraduate teaching in science, technology, engineering, and mathematics; The National Academies Press: Washington, DC, 2003. Invention and Impact: Building Excellence in Undergraduate Science, Technology, Engineering and Mathematics (STEM) Education; American Association for the Advancement of Science: Washington, DC, 2004. National Leadership Council for Liberal Education and America’s Promise. College Learning for the New Global Century; Association of American Colleges and Universities: Washington, DC, 2007. Graham, M. J.; Frederick, J.; Byars-Winston, A.; Hunter, A. B.; Handelsman, J. Increasing Persistence of College Students in STEM. Science 2013, 341, 1455–1456. Kuh, G. High-Impact Educational Practices: What They Are, Who Has Access to Them, and Why They Matter; Association of American Colleges & Universities: Washington, DC, 2008. Board on Science Education. Evidence of Promising Practices in Undergraduate Science, Technology, Engineering, and Math (STEM) Education. http://sites.nationalacademies.org/dbasse/bose/dbasse_080106? _ga=2.38701984.2082328722.1509727018-1946991266.1509727018 (accessed January 9, 2009). Association of American Colleges and Universities. Diversity, Equity, & Inclusive Excellence. https://www.aacu.org/resources/diversity-equity-andinclusive-excellence (accessed August 1, 2016). 54

20. Handelsman, J.; Pfund, C.; Lauffer, S. M.; Pribbenow, C. M. Entering Mentoring: A Seminar to Train a New Generation of Scientists; University of Wisconsin Press: Madison, WI, 2009. 21. Packard, B. W.-L. Successful STEM Mentoring Initiatives for Underrepresented Students: A Research-Based Guide for Faculty and Administrators; Stylus: Sterling, VA, 2016. 22. DiBartolo, P. M.; Gregg-Jolly, L.; Swartz, J.; Davidson, C.; Gross, D.; Manduca, C.; McDaris, J. Supporting STEM Success in a Liberal Arts Context. https://serc.carleton.edu/108221 (accessed August 1, 2016). 23. DiBartolo, P. M.; Gregg-Jolly, L.; Gross, D.; Manduca, C. A.; Iverson, E.; Cooke, D. B.; Davis, G. K.; Davidson, C.; Hertz, P. E.; Hibbard, L.; Ireland, S. K.; Mader, C.; Pai, A.; Raps, S.; Siwicki, K.; Swartz, J. E. Principles and Practices Fostering Inclusive Excellence: Lessons from the Howard Hughes Medical Institute’s Capstone Institutions. CBE-Life Sci. Edu. 2017, 15, 44; DOI: 10.1187/cbe.16-01-0028. 24. Lopatto, D. Appendix B Commissioned Paper: The Consortium as Experiment. In Integrating Discovery-Based Research into the Undergraduate Curriculum: Report of a Convocation; The National Academies Press: Washington, DC, 2015; pp 101−122. 25. Howard Hughes Medical Institute. 47 Colleges to Participate in $50 Million Science Education Initiative. http://www.hhmi.org/news/47-collegesparticipate-50-million-science-education-initiative (accessed November 30, 2015). 26. Fiegener, M. K.; Proudfoot, S. L. Baccalaureate Origins of U.S.-trained S&E Doctorate Recipients; National Science Foundation: Arlington, VA, 2013. 27. Bhattacharya, B.; Hansen, D. Implementing a Summer STEM Bridge Program. PeerReview 2015, 17, 19–20. 28. Treisman, P. M. A Study of the Mathematics Performance of Black Students at the University of California, Berkeley. Ph.D. Thesis, The University of California, Berkeley, December 1985. 29. Steele, C. M.; Aronson, J. Stereotype threat and the intellectual test performance of African Americans. J. Pers. Soc. Psychol. 1995, 69, 797–811. 30. Cohen, G. L.; Garcia, J.; Apfel, N.; Master, A. Reducing the racial achievement gap: A Social psychological intervention. Science 2006, 313, 1308–1310. 31. Dweck, C. S.; Leggett, E. L. A Social-cognitive approach to motivation and personality. Psychol. Rev. 1988, 95, 256–273. 32. Aronson, J.; Fried, C. B.; Good, C. Reducing the effects of stereotype threat on African American college students by shaping their views of intelligence. J. Exp. Soc. Psychol. 2002, 38, 113–125. 33. Nadal, K. L.; Davidoff, K. C.; Davis, L. S.; Wong, Y.; Marshall, D.; McKenzie, V. A qualitative approach to intersectional microaggressions: Understanding influences of race, ethnicity, gender, sexuality and religion. Qual. Psychol. 2015, 2, 147–163. 34. Garcia, A. M.; Hurtado, S. Predicting Latina/o STEM persistence at HSI and non-HSIs; American Educational Research Association: Vancouver, 2011. 55

35. Solórzano, D. G.; Ceja, M.; Yosso, T. J. Critical race theory, racial microaggressions, and campus racial climate: The experiences of African American college students. J. Negro Educ. 2000, 69, 60–73. 36. Estrada, M.; Burnett, M.; Campbell, A. G.; Campbell, P. B.; Denetclaw, W. F.; Gutiérrez, C. G.; Hurtado, S.; John, G. H.; Matsui, J.; McGee, R.; Okpodu, C. M.; Robinson, T. J.; Summers, M. F.; Werner-Washburne, M.; Zavala, M. Improving Underrepresented Minority Student Persistence in STEM. CBE-Life Sci. Educ. 2016, 15, 1–10. 37. Lewin, K. Action Research and Minority Problems. J. Soc. Iss. 1946, 2, 34–46. 38. Burnes, B. Kurt Lewin and the Planned Approach to Change: A Re-appraisal. J. Manage. Studies 2004, 41, 0022–2380. 39. Henderson, C.; Beach, A.; Finkelstein, N. Facilitating Change in Undergraduate STEM Instructional Practices: An Analytic Review of the Literature. J. Res. Sci. Teach. 2011, 48, 952–984. 40. Barefoot, B. O.; Griffin, B. Q.; Koch, A. K. Enhancing Student Success and Retention throughout Undergraduate Education: A National Survey; John N. Gardner Institute for Excellence in Undergraduate Education: Brevard, NC, 2012. 41. Smith, K. A.; Sheppard, S. D.; Johnson, D. W.; Johnson, R. T. Pedagogies of Engagement: Classroom-Based Practices. J. Eng. Educ. 2005, 94, 87–101. 42. Singer, S. R., Nielsen, N. R., Schweingruber, H. A., Eds. Discipline-Based Education Research: Understanding and Improving Learning in Undergraduate Science and Engineering; National Academies Press: Washington, DC, 2012; pp 119−139. 43. Tobias, S. They’re Not Dumb, They’re Different: Stalking the Second Tier; Research Corporation: Tucson, AZ, 1990. 44. Freeman, S.; Eddy, S. L.; McDonough, M.; Smith, M. K.; Okoroafor, N.; Jordt, H.; Wenderoth, M. P. Active Learning Increases Student Performance in Science, Engineering, and Mathematics. Proc. Natl. Acad. Sci. 2014, 111, 8410–8415. 45. Laws, P. Calculus-Based Physics without Lectures. Phys. Today 1991, 44, 24–31. 46. Schneider, M. B. A Lab-Based, Lecture-Free General Physics Course. BAPS 1997, 42, 999. 47. Lindgren, C. Teaching by Doing: Turning a Biology Curriculum Upside Down. http://www.skeptic.com/eskeptic/09-11-18/#feature (accessed November 18, 2001). 48. Lindgren, C. Teaching Matters: Turning the Teaching of Sciences Upside Down. http://chronicle.com/article/Teaching-Matters-Turning-the/65132/ (accessed April 18, 2010). 49. Brewer, C., Smith, D., Eds. Vision and Change in Undergraduate Biology Education: A Call to Action; American Association for the Advancement of Science: Washington, DC, 2011. 50. Rodenbusch, S. E.; Hernandez, P. R.; Simmons, S. L.; Dolan, E. L. Early Engagement in Course-Based Research Increases Graduation Rates and 56

51.

52.

53.

54. 55.

56.

57.

58.

59. 60. 61.

62.

63.

Completion of Science, Engineering, and Mathematics Degrees. CBE-Life Sci. Edu. 2016, 15, 20. Gandara, P.; Maxwell-Jolly, J. Priming the Pump: Strategies for Increasing the Achievement of Under-represented Minority Undergraduates; College Board: New York, NY, 1999. Hunter, A.; Laursen, S.; Seymour, E. Becoming a Scientist: the Role of Undergraduate Research in Students’ Cognitive, Personal, and Professional Development. Sci. Educ. 2007, 91, 36–74. Lopatto, D. Science in Solution: the Impact of Undergraduate Research on Student Learning; CUR and the Research Corporation for Science Advancement: Washington, DC, 2009. Russell, S. H.; Hancock, M. P.; McCullough, J. Benefits of Undergraduate Research Experiences. Science 2007, 316, 548–549. Eagan, M. K.; Hurtado, S.; Chang, M. J.; Garcia, G. A.; Herrera, F. A.; Garibay, J. C. Making a difference in science education: the impact of undergraduate research programs. Am. Educ. Res. J. 2013, 50, 683–713. Jones, M. T.; Barlow, A. E. L.; Villarejo, M. Importance of Undergraduate Research for Minority Persistence and Achievement in Biology. J. Higher Educ. 2010, 81, 82–115. Nagda, B. A.; Gregerman, S. R.; Jonides, J.; von Hippel, W.; Lerner, J. S. Undergraduate student-faculty research partnerships affect student retention. Rev. Higher Educ. 1998, 22, 55–72. Weinman, J.; Jensen, D.; Lopatto, D. Teaching Computing as Science in a Research Experience. In Proceedings of the 46th ACM Technical Symposium on Computer Science Education; Association for Computing Machinery: Kansas City, MO, 2015; pp 24−29. Sullivan, C.; Hanzawa, F. M. Mentoring Program for Introductory Biology at Grinnell College. CUR Q. 1992, 12, 32–36. Mahlab, M. Who Benefits? Peer Mentors at Grinnell College. CUR Q. 2010, 31, 7–10. Gregg-Jolly, L. A.; Swartz, J. E.; Iverson, E.; Stern, J.; Brown, N.; Lopatto, D. Situating Second-Year Success: Understanding Second-Year STEM Experiences at a Liberal Arts College. CBE-Life Sci. Educ. 2016, 15, 43. Science Education Resource Center. Broadening Access to the Sciences. http://serc.carleton.edu/broadening_access/index.html (accessed August 1, 2016). Science Education Resource Center. LSAMP IINSPIRE Project: Teaching and Program Resources. http://serc.carleton.edu/lsamp/index.html (accessed August 1, 2016).

57

Chapter 4

Evaluation of Effects of an Intervention Aimed at Broadening Participation in STEM while Conveying Science Content Heather Perkins,1,* Mary Wyer,1 and Jeffrey N. Schinske2 1Dept.

of Psychology, North Carolina State University, Raleigh, North Carolina 27695, United States 2Biological & Health Sciences, Foothill College, Los Altos Hills, California 94022, United States *E-mail: [email protected]

In an earlier paper, we reported results from a study of an innovative classroom intervention on community college students’ perceptions of scientists (Schinske, J., Perkins, H., Snyder, A., Wyer, M., CBE Life Sci. Educ., 2016, 15, ar47). Dubbed “Scientist Spotlights,” the assignments exposed students to images of scientists who were not stereotypical asocial white males in lab coats (Hashimoto, T., Karasawa, K., Recent Advances in Natural Computing, 2015, 9, 57–67; McGee, E. O., Am. Educ. Res. J., 2016, 53, 1626–1662), thus diversifying their images of science and scientists, and boosting their identification with and engagement in STEM. The qualitative analyses reported in Study One suggested that exposure to the Scientist Spotlights reduced stereotypic images and improved students’ ability to relate to scientists when compared to a control group. This second paper reports on our quantitative study of the intervention’s effects, analyzing preand post-test responses to the 22-item Stereotypes of Scientists Scale (4S) (Wyer, M., Schneider, J., Nassar-McMillan, S., Oliver-Hoyo, M., Int. J. Gender, Sci. Technol., 2010, 2, 382-415). Two-hundred and twelve students, separated into control and intervention groups, were surveyed at pre- and post-test, and confirmatory factor analysis (CFA) found that the scales previously demonstrated structure was still valid for

© 2018 American Chemical Society

pre-test (CFI = .94, TLI = .92, RMSEA = .06, SRMR = .06) and control students (CFI = .93, TLI = .91, RMSEA = .09, SRMR = .06), but did not hold up for intervention students (CFI = .88, TLI = .84, RMSEA = .11, SRMR = .07). Analyses of these findings suggest that post-test intervention students hold a more integrated vision of scientists’ professional and interpersonal skills, and that Scientist Spotlights can be used to disrupt students’ stereotypes of scientists.

Consider the case of Gabriela, a Communications major who encounters a flyer urging her to learn more about a career in STEM. The flyer includes messages about the financial success and low unemployment within the STEM fields, and Gabriela experiences a pang of doubt and insecurity regarding her decision to major in Communications. In response to this uncertainty, Gabriela maintains her equilibrium by negatively evaluating scientists as brilliant but asocial, particularly in comparison to herself as a good student who is primarily interested in working with others. Thus, scientists are superior in terms of intelligence (competence and earning potential) but inferior in terms of sociability (warmth). Gabriela relies on this stereotype when assessing the invitation to become more involved in STEM and decides that, because she highly values working with others, she is unsuited to a STEM profession. Furthermore, if she counters difficulty in her STEM courses later, Gabriela will interpret this as further evidence that she “isn’t a STEM person” and will disengage, preferring to spend her time and effort on challenges that she believes are relevant to her chosen identity.

Scientist Spotlights to Challenge Stereotypes in Diverse Classrooms Gabriela’s case exemplifies the key role that stereotypes play when students are evaluating their place in STEM. Issues of under-representation and inadequate recruitment in STEM are well known; while the number of STEM and STEM-adjacent positions in the United States is increasing, interest in STEM is not following suit (1–3). More worryingly, many students with strong STEM interest at the beginning of their academic careers find themselves driven from the field, and many of these students hail from groups that have been traditionally marginalized in science and academia (4–7). It is frequently theorized that, as in Gabriela’s case, stereotypes of scientists play a role in these shortages and alienations, with stereotypes in two spheres (dispositional traits and demographic characteristics) operating concurrently to suppress interest and isolate marginalized students. 60

More specifically, stereotypes of scientists and other STEM professionals frequently focus on a perceived dichotomy between warmth and competence, in which members of a competitive outgroup are acknowledged to be competent, or intelligent, but criticized and scorned for their lack of warmth, or interpersonal skills (8). These personality-based stereotypes overlap with stereotypes of race, gender, and academic achievement, and result in portrayals of scientists that are largely White or East Asian, male, and economically advantaged, thus communicating a lack of belonging to students who do not fit this profile (9–11). Previous research indicates that these stereotypes of scientists’ personal traits can alienate students and lead them to disengage from STEM settings that might otherwise capture their interest (12). Similarly, stereotypes of natural intelligence (often referred to as a culture of brilliance, as opposed to a culture of merit) and representation often go hand-in-hand; fields in which practitioners are stereotyped as being more innately intelligent also enroll fewer women and people of color (13). In response to these issues, the Scientist Spotlights intervention was developed for use in STEM classrooms (14). This intervention is designed to supplement regular course instruction – students learn about counter-stereotypical examples of scientists through out-of-class reading or media and turn in their reflections as homework. The focus of these assignments is both on the work done by the scientists as well as their personal stories, thus serving a dual purpose; students learn about ‘real-world’ applications of course material while being presented with diverse narratives about scientists and scientific work. For example, when learning about neuron signaling, students are asked to read about Ben Barres, a transgender scientist who has spoken publicly about both his research and his transition. To complete the assignment, students answer questions about neuron signaling and about Dr. Barres and their preconceptions of scientists. Based in theories of possible selves and identity-based motivation, this intervention is theorized to work by exposing students to scientists whom they can more easily relate, thus inviting them to imagine a place for themselves in STEM (15, 16). It emphasizes that scientists are diverse in terms of roles, goals, values, personalities, and backgrounds, and provides opportunity to identify with and potentially envision themselves as scientists (i.e., possible selves and role-congruity) (17, 18). Consistent with identity-based motivation (IBM) and the stereotype inoculation model, this ability to identify with scientists when engaged in science learning boosts self-efficacy and connectedness, thus making the work identity-congruent, or important to the students’ future goals (19–21). As a result, when students encounter difficulty in class, they are more likely to interpret the content as important and are motivated to buckle-down and engage seriously with the material. In contrast, students who view the work as identity-incongruent are more likely to conclude that they don’t belong in the setting and that they should disengage, negatively impacting their class performance as well as their STEM interest and science identity. Previous work has demonstrated that the Scientist Spotlights intervention is effective in shifting student’s stereotypes of scientists, and that students exposed to the intervention (as compared to those in a control group) endorse fewer stereotypes when discussing scientists and describe them as more relatable (14). 61

However, we sought to build on this existing work by developing a quantitative assessment that can be used to evaluate shifts in students’ stereotypes, an endeavor which presented many challenges. This chapter discusses the instrument and analysis used to measure the effectiveness of the Scientist Spotlight intervention and provides advice and tools for those who wish to utilize this intervention to challenge stereotypes of scientists.

Measuring the Disruption of Stereotypes According to the psychology literature, stereotypes can be viewed as sets of ideas, or schemas, that enable an individual to maintain their self-esteem and identity in the face of upward or downward comparisons (8). They act as cognitive and cultural heuristics that communicate patterns to reduce resource strain and enable accurate snap judgements (22–25). However, they can also act as barriers, particularly when students use them to steer their decisions, as was demonstrated by Gabriela at the beginning of this chapter (26, 27). Gabriela’s views of scientists’ superiority in terms of competence and earning potential, but inferiority in terms of warmth, are consistent with the Stereotype Content Model and common portrayals of scientists (28). Her decision to disengage from STEM due to a judgment that she is not a “STEM person” is further consistent with possible selves and IBM (23). However, by focusing on a single student, The example of Gabriela also masks some of the complexities inherent in assessing stereotype change. It’s important to remember that stereotypes are culturally and socially situated phenomena, and while they are frequently shared (i.e., groups share similar stereotypes about other groups) they also vary between groups and across individuals. Although there are consistent trends in how scientists are portrayed in the U.S. and around the world – generally as socially inept geniuses – it would be a mistake to assume that all students share this stereotype, or that they engage with it identically. For instance, one student may view the intelligence of scientists as a positive and desirable characteristic, while another may see it as uselessly cerebral and unsuited to their goals or environments (29). An intervention that highlights the esteem and importance of scientific work may thus have a similar effect on both students (an increase in how intelligent scientists are perceived to be) with opposite consequences (the first student’s STEM interest increases while the second student’s decreases). Similarly, an intervention that highlights how scientists interact with students, peers, and the public during their work may increase STEM interest in the Communications student introduced above, while alienating a shy student who is reassured by the stereotype of a solitary scientist at work. As a result, interventions that alter students’ stereotypes of scientists – a seemingly common-sense solution – are potentially derailed due to the difficulty in evaluating interventions’ complicated effects. 62

Confirmatory Factor Analysis (CFA) To Assess Stereotype Disruption In response to this challenge, we propose the use of Confirmatory Factor Analysis (CFA) and the Stereotypes of Scientists Scale (4S) to evaluate shifts in students’ perceptions of scientists following counter-stereotype interventions, such as Scientist Spotlights. Factor analysis is used to investigate the relationships between sets of items (also referred to as indicators). Unlike regular bivariate correlations, factor analyses identify the relationships between more than two items, and in doing so identify the common variance shared by the set of items (also referred to as ‘factors’ or ‘latent variables’). The two types of factor analysis are exploratory and confirmatory, the latter of which is used to test hypotheses and is featured in this paper. For a more detailed overview of exploratory factor analysis, we recommend the article by Yong & Pearce (30), and for confirmatory factor analysis, we recommend the comprehensive book by Brown (31). The 4S, developed from qualitative and quantitative work with undergraduate students, is a 22-item scale that assesses student opinions of common stereotypes of scientists (32, 33) (see Table 1 for a full list of items). In line with the SCM, these stereotypes are not inherently positive or negative, but are networks of ideas that promote fixed and immobile relationships between individuals and their social environments. These preconceptions are arranged on axes of warmth and competence (also referred to as interpersonal competence and professional competence in the literature), but there is no ‘cut-off’ indicating that a student’s opinions are problematic and in need of adjusting. Indeed, this is not how we propose to use the scale at all – instead, the focus is on measuring the disruption to students’ stereotypes following exposure to an intervention. Because testing this disruption presents a challenge to traditional statistical methods, which often rely on mean differences between groups or linear relationships between predictors and outcomes (e.g., ANOVA and multiple regression), we suggest using CFA to evaluate the effectiveness of these disruptive interventions. This technique is recommended because it allows us to test the relationships between indicators and latent variables, and to explore whether those relationships are consistent across groups (e.g., measurement invariance (31)). Typically, CFA is practiced in a scale development context to ensure the absence of test bias or the differential functioning of measure items. For instance, CFA has been used to assess the functioning of a depression scale in patients with cancer as compared to those without, to test whether the symptoms of cancer and treatment foil the psychometrics of scale items (e.g., an item such as “I was aware of the dryness of my mouth” may no longer be an accurate measure of anxiety if cancer patients routinely experience dry mouth because of their medication) (34). With CFA, the test is not for differences in mean scores, such as whether one group scores higher or lower on a subscale, but of the proposed factor structure (also referred to as the proposed model) and its fit. The factor structure is determined by the pattern of responses across items, such as a rigid dichotomy between scientists’ warmth and professional competence. Finding that these patterns have been interrupted in the intervention group and not in the control 63

group demonstrates that stereotypes have been disrupted by the intervention, and that students’ opinions about diversity in STEM have been changed.

Table 1. Stereotypes of Scientists Scale (4S)a Professional Competencies Know a lot about the latest discoveries The ones who know how equipment works Careful with expensive instruments Competitive Independent Work oriented Technically competent Competent Self-confident Highly focused Able to learn to use new equipment quickly Especially intelligent Logical Interpersonal Competencies Have fun with colleagues at work Maintain friendships with colleagues in other departments Do not have a lot of friends b Out of touch with what is happening in the world b Have happy marriages Cooperative Family oriented Insecure b Collaborative a

Note: Item stems read, "When I think about scientists I think they are…." reverse-coded item.

b

Indicates

Evaluating Scientist Spotlights in Diverse Classrooms Participants The students involved in this study were recruited from a minority-serving community college on the West Coast, with a population that is very diverse, 64

often challengingly so when it comes to traditional statistical tests. The total student population is over 20,000, 48% female and 50% male. Approximately 37% of students are Asian, 27% are Hispanic, and 20% are White. Students who participated in our control classes or completed the intervention were asked to provide more details about their demographics and showed a slightly different profile that the school as a whole (see Table 2 for more information).

Table 2. Sample Demographics Full Sample N

Characteristic

%

503

Total

Matched Sample N

%

212

Control Group

158

31%

106

50%

Intervention Group

345

69%

106

50%

Male

180

36%

74

35%

Female

311

62%

138

65%

Trans, non-binary, or agender

12

2%

0

0%

Race/Ethnicity: Well-Represented

149

30%

54

25%

East Asian (e.g., Chinese)

76

15%

35

17%

White (incl. European)

73

15%

19

9%

345

69%

158

75%

Latinx

104

21%

46

22%

Hispanic

26

5%

20

9%

Black

8

2%

4

2%

Southeast Asian (e.g.Southeast Asian (e.g., Filipino)

98

19%

39

18%

South Asian (e.g.South Asian (e.g., Indian)

19

4%

8

4%

Middle Eastern (e.g., Persian)

11

2%

7

3%

Multiethnic

79

16%

34

16%

9

2%

0

0%

Race/Ethnicity: Under-Represented

Race/Ethnicity: Not Available

To best accommodate and utilize the diversity in our sample, we coded participants’ races according to racial representation within the field. Previous literature indicates that White or East Asian students are over-represented in STEM relative to their rates within the population, and that Hispanic/Latino, Black, Filipino/a, and Native American students are under-represented (i.e., the ratio of representation (35)). Other racial and ethnic backgrounds (e.g. 65

Middle Eastern, South Asian, and Multiracial) are frequently overlooked and understudied. Accordingly, we coded students who reported a traditionally stereotypical background as ‘well-represented’ (n = 149), and students from negatively stereotyped, overlooked, and multiracial/ethnic backgrounds as ‘under-represented’ (n = 345). These codes were used primarily when case-matching to balance the control and intervention groups; due to the resulting small sample, it was not possible to test for control-by-representation interactions (see Future Work for more discussion of this point). Using a quasi-experimental design, 3 classrooms were selected to act as a control group, while 6 classrooms received the full Scientist Spotlights intervention (see intervention description at beginning of chapter). Aside from the omission of the intervention, the classrooms were largely identical. All were introductory biology courses, and all but one of the classes were taught by the same instructor using the same course materials. Data collection took place from 2014 to 2017; students completed the intervention as part of their course curriculum and were invited to participate in the survey evaluation for extra credit. Measures Across all administrations of the survey, students were asked to complete the 22-item 4S, reporting their opinions about scientists and themselves (see Table 1 for a full list of items). Students also completed measures that asked about their academic self-efficacy and mindsets, their interest, confidence, and past experiences in science, their opinions about equality in science and education, and their racial/ethnic and gender identities, although this data is not analyzed in this study. Analyses: Data Screening Confirmatory Factor Analysis (CFA) was identified as the best method for testing how this intervention disrupted students’ stereotypes of scientists. Before proceeding with analyses, the data was screened to assure it met the assumptions of the test (see Appendix 1 for more details about data screening). Additionally, although participants were prompted to respond to all questions, there were still incidents of skipped items. Participants who did not reach the end of the survey or who skipped more than one item were dropped; for those who skipped only a single item, average imputation was used to replace the remaining missing values. An additional concern with CFA is the effect of unequal sample sizes. As noted above, the intervention group was nearly twice the size of the control group, due to the quasi-experimental nature of the study design (36). Since CFA relies heavily on Chi-square tests, which are very sensitive when sample sizes are large, we used case-matching to balance the number of participants in the control and experimental conditions (using the optmatch and RItools packages in R (37)). Participants were matched according to race/ethnicity (using the codes of wellrepresented and under-represented, as introduced above) and gender, resulting in a final post-test sample of 212, (control n = 106, intervention n = 106) (see Table 2 for full participant demographics). 66

Analyses: Model Evaluation and Testing Here we discuss the planned analyses that produced the results in the section below. For those unfamiliar with factor analysis, we recommend the use of online reference materials for guidance in replicating these analyses (38). To test for changes in the factor structure using CFA, an initial model needed to be proposed and justified. Previous research uncovered a 2-factor structure for the 4S scale (see Table 1 for a full list of items and factors). As previous analyses were exploratory rather than confirmatory, an E/CFA (an exploratory factor analysis within the CFA framework that produces fit statistics and modification indices) was run using previously collected data. Based on these analyses, a good fitting 16-item, 2factor structure with one cross-loading item and eleven inter-correlations (due to similarity in items, i.e., method effects) was proposed and tested on the pre-test data (n = 212), and the post-test control (n = 106) and intervention (n = 106) data using the lavaan package in R (39). The model fit was also evaluated using CFA in multiple groups (31). First, a measurement invariance test was run (using the SEMtools package in R (40)), evaluating four models with increasing levels of constraint (see Figures A1 and A2 for an illustration of the initial model and the measurement invariance process). The first model tests that the control and experimental groups have similar factor structures (configural invariance); the second model uses a harder test, in which the groups have similar factor structures and factor loadings (weak invariance). Similarly, the third and fourth models add constraints to the intercepts and residual variances, respectively (strong invariance and strict invariance). Significant results indicate that the additional level of constraint produced a model that is poorerfitting than the previous one, and thus that there is some difference between the two groups powering the decrease in fit. Removing the source of this difference and re-running the test should produce non-significant results, i.e., a model that is stable across the two groups. In using this test, we can locate the items that are responded to differently by the control and intervention conditions, and thus can identify specifically where students’ stereotypes of scientists are changing due to the intervention.

Results: Comparing Stereotypes of Control and Intervention Students Stereotypes of Scientists Shifted for Intervention Students The first step in assessing whether the intervention successfully disrupted students’ stereotypes was a CFA with the model developed in previous work (see Figure A1). Testing this model on the full sample at pre-test demonstrated adequate fit, CFI = .94, TLI = .92, RMSEA = .06, SRMR = .06 (41, 42). All the items loaded as predicted on either the interpersonal competence or professional competence factors (excepting item 19, “They are collaborative,” which loaded on both factors). The residual correlations observed in previous research, likely due to method effects caused by similar wording in survey items, also held (see Figure A1). 67

The model continued to fit for the control group at post-test, CFI = .93, TLI = .91, RMSEA = .09, SRMR = .06. The slight decrease in fit was likely due to the loss in power; the full sample (n = 212) was tested at pre-test, but the control sample had only half as many participants. The factor loadings also performed well (see Table A3), with the single exception of item 19, which was no longer cross-loading with the professional competence items. There were indications of some strain around the indicator covariances (i.e., predicted residual correlations due to method effects), also likely due to the smaller sample size, but the overall fit was still adequate. The same could not be said of the model when tested with the intervention post-test group, CFI = .88, TLI = .84, RMSEA = .11, SRMR = .07. The decrease in overall fit pushes the model past nearly all of the cut-offs recommended in the literature (42, 43). Despite this, the factor structure held up, with only item 19 behaving contrary to predictions (i.e., failing to cross-load, as with the control sample; see Table A3). The indicator covariances and residuals were also similar across the two groups (see Table A3). Although these scores are not radically divergent, they demonstrate that the model fits well for pre-test and control students, but it is no longer adequate for intervention students. Since item 19 was not cross-loading for either group of students, it was removed from the professional competence factor, to no major effect on the overall fit (control: CFI = .93, TLI = .91, RMSEA = .08, SRMR = .06; intervention: CFI = .88, TLI = .84, RMSEA = .11, SRMR = .07). For a full list of factor loadings, variances, and co-variances of the three models, see Figures A1 and A2, and Tables A1, A2, and A3.

Source of the Shift for Intervention Students Following this basic demonstration of the fit differences across conditions, tests of measurement invariance were used to demonstrate the statistical significance of the differences between groups, and to identify which items were causing the variation. As was mentioned previously, this process of testing and refinement increasingly restricts the variance allowed between the two groups. In scale development, this is done to demonstrate that the scale performs similarly across two groups; here, this test shows that the scale does not perform similarly across the two groups, because the intervention group’s ideas about scientists have been changed by the intervention. Overall, this test indicated that the model was not invariant, particularly when the intercepts and residual variances were constrained to be equal across the two groups (see Table 3 for the full list of values). Typically, when this occurs during scale development, the ‘misbehaving’ items are removed and the model re-tested until the fit scores return to the desired levels. If, following these modifications, the increasingly constrained models do not show significant differences in fit, the remaining items are considered invariant and can be used freely within the two groups. Here, we also pulled the items and re-tested, in order to identify which indicators caused the variance and thus which items best captured student’s changing stereotypes. 68

Table 3. Testing measurement invariance: Comparing Chi-square difference scores Model

χ2

χ2diff

dfdiff

p

CFI

RMSEA

0.91

0.10

Original Model (Figure A1) Model 1: Configural Invariance

359.53

Model 2: Weak Invariance

377.74

18.21

14

0.197

0.90

0.09

Model 3: Strong Invariance

401.16

23.42

14

0.054

0.90

0.09

Model 4: Strict Invariance

432.11

30.94

16

0.014

0.89

0.09

0.91

0.10

Modified Model Model 1: Configural Invariance

318.80

Model 2: Weak Invariance

330.71

11.91

13

0.535

0.91

0.09

Model 3: Strong Invariance

349.57

18.85

13

0.128

0.90

0.09

Model 4: Strict Invariance

379.64

30.07

15

0.012

0.89

0.09

0.93

0.10

Modified Model (Figure A3) Model 1: Configural Invariance

154.60

Model 2: Weak Invariance

158.97

4.37

9

0.886

0.93

0.09

Model 3: Strong Invariance

173.17

14.20

9

0.115

0.93

0.09

Model 4: Strict Invariance

191.06

17.90

11

0.084

0.92

0.09

As the fit decreased significantly when transitioning from Model 2 (Weak Invariance; factor structure and loadings constrained) to Model 3 (Strong Invariance; intercept constraints added) we calculated the absolute value of the differences between the z-scores of the intercept estimates when they were free to vary. Eight of the sixteen items differed by at least one standard deviation across the two groups, with item 19 on the interpersonal competence factor varying the most. This item was removed from the model and the test of measurement invariance run again, with Model 3 no longer demonstrating a significant decrease in fit. However, Model 4 (Strict Invariance: residual variance constraints added) still demonstrated a significant decrease in fit. Comparing the differences in item variance across the two groups showed only one item that differed by most of a standard deviation (item 2, “They have fun with colleagues at work”). Removing this item improved the fit of the model, but it was still significantly worse than Model 3 (χ2 (14) = 32.28, p = .003. Removing item 23 (“They are logical”) also failed to eliminate the poor fit, χ2 (13) = 25.94, p = .02. The next item with the largest difference across groups was item 3 (“They maintain friendships with colleagues in other departments”), one of the four remaining indicators for the interpersonal competence factor. Removing this item did see a small improvement in fit, χ2 (12) = 22.05, p = .03, but also produced 69

Heywood cases in the factor loadings (Heywood cases, or abnormal factor loadings, are theoretically impossible loadings that sometimes appear in factor analysis and indicate severe issues with the proposed model) (44). Rather than risk destabilizing the model completely, item 3 was retained and the next item on the list, item 8 (“They are careful with expensive instruments”) and then item 7 (“They are the ones who know how equipment works”) were removed instead (χ2 (12) = 22.31, p = .03; χ2 (11) = 17.90, p = .08). This reduced the decrease in fit to non-significance, producing a final model that fit both groups well, CFI = .93, TLI = .92, RMSEA = .09. SRMR = .07 (see Figure A3).

Discussion: Control & Intervention Students Perceived Scientists Differently To summarize these findings, a preliminary evaluation of the model for the pre-test, control group, and intervention group data demonstrated that the model fits the pre-test and control students well but does not hold up when tested on intervention students. To refine these findings, we used invariance testing, which compares increasingly constrained models across two groups and tests for significant differences. Models 3 and 4, in which the intercepts and residual variance were constrained, showed a significant decrease in fit, a decrease that was anchored in responses to items 2, 3, 7, 8, 19, and 23. In order to explore these findings, we will discuss these items and how they relate to the Scientist Spotlights intervention, and how instructors and interventionists can make use of these results in their own work. Removing items 2, 7, 8, 19, and 23 from the factor structure produced a stable, invariant model that worked equally well for control and intervention students. Although it was a potentially significant source of variance, removing item 3 seriously destabilized the model (likely due to an inadequate number of indicators remaining in the interpersonal competence factor (31)) and so it was maintained. These results suggest that the stereotypes assessed by these items were the ones best addressed by the current iteration of the intervention. There are trends among the items that showed the most non-invariance between the two groups. The first, item 19, asks participants to rate how much they agree that, “[When I think about scientists, I think…] they are collaborative”. Similarly, item 2, and potentially Item 3, are correlated items that describe scientists as “having fun with colleagues at work” and “maintaining friendships with colleagues in other departments”. Taken together, these three items portray science as social and collaborative, and reject the conventional stereotype of the solitary scientist (45, 46). Items 7 and 8 are also correlated and assess stereotypes of technical skills and care with expensive instruments, an object-oriented focus that is central to stereotypes of scientists but alienating for some students (47–49). The final item removed from the model, item 23, is also central to the stereotype of scientists as cerebral and rationalistic, (i.e., “They are logical”). Variance in these three items suggest that something about this central aspect of the stereotype is challenged by the intervention, although precisely what is unclear. 70

Previously, a qualitative evaluation of the intervention saw students’ descriptions of scientists shift to counter-stereotypical narratives that emphasized scientists’ individuality and humanistic focus (14). Taken with our current results, these changes suggest the intervention challenges the conventional stereotype of scientists as asocial and object-oriented. They also suggest the potential emergence of a more communal view of science, one aligned with other interventions that can prevent women from disengaging from STEM (i.e., role congruity) and can help students develop critical science agency (50, 51). Critical science agency, which also extends to other STEM fields, emphasizes the humanistic aspects of scientific work to enhance interest and empower students from marginalized backgrounds (52). Although this intervention was not developed with a lens towards critical science agency or role congruity, this overlap suggests that it may be a particularly potent pathway for challenging stereotypes about scientists and engaging students in STEM. Limitations and Future Work Although these results are intriguing, the methodological approach is still novel and results must be interpreted with care, as further testing using this approach may shed additional light on these findings or call for re-evaluation. The precise meaning of these changes is also difficult to assess at this point. The current study operates under the theory that any changes to scientist stereotypes are positive, as they indicate flexibility in the face of rigidity. The primary challenge in communicating and representing these findings lies in the novelty of the methodological approach; as was mentioned previously, a simple test of mean differences cannot convey the changes that occur in the relationships between items (i.e., the correlations), only the change in individual scores (i.e., the means). As a result, these findings represent the first step in evaluating this intervention. Specifically, they demonstrate that it changes students’ stereotypes (their cognition about scientists), but the impact of these changes on STEM interest and participation (their attitudes about scientists) remains to be demonstrated. Furthermore, the inability to explore the impact of the intervention across demographic categories, such as the well-represented/under-represented coding scheme introduced previously, is also a limitation. In addition to exploring the effects of the intervention across student groups (e.g., race/ethnicity and gender), future studies need to confirm the effects that these changes have on students’ STEM interest, class performance, and science identity, before this intervention’s effectiveness can be fully understood. Future studies could also explore the relationship between students’ self-perceptions and how they relate to their changing perception of scientists, further exploring the construct of identity congruence and fit, and how it relates to other constructs of interest. Conclusions & Recommendations Recall from the beginning of this chapter how stereotypes led Gabriela to disengage from STEM and negatively evaluate scientists. What if instead Gabriela had encountered the Scientist Spotlights intervention, in which stereotypes were 71

directly challenged and diversity in STEM highlighted? Rather than withdraw, Gabriela might have been motivated to engage with her STEM courses more deeply, perceiving them as applicable and beneficial to her current goals, or even to begin visualizing herself as a potential STEM major. Our analyses suggest that Scientist Spotlights are a valuable, versatile, and easily deployed intervention with a demonstrable effect on students’ stereotypes, thus better enabling them to identify with scientists and see themselves in the field. While assessing the change in stereotypes presented a challenge to traditional modes of measurement, with CFA we were able to duplicate previous results uncovered using qualitative techniques (14). The results of these analyses provide indications of which stereotypes were disrupted, and thus offers guidance for future instructors and interventionists seeking to challenge stereotypes and broaden STEM interest in their students. To further the development of this methodological approach and use of the intervention, the R code and intervention information are available online for those who are interested (38). Based on our results here, we recommend the use of the Scientist Spotlights to challenge and disrupt stereotypes of scientists in the classroom, and the 4S and CFA to assess the effects of these interventions. We also recommend enhanced collaboration between instructors and administrators, departments and disciplines, and the social sciences and STEM educators to continue expanding, refining, and evaluation of stereotype-busting interventions. By extending an invitation to others interested in this area of work, we hope to build a collaborative network that challenges stereotypes and fosters diversity in science and scientific work alike.

Appendix 1: Data Screening Data was tested to confirm it met the sample size and normality assumptions of maximum likelihood estimate (MLE) (the estimation method most widely used in CFA; Brown, 2015). Ensuring one’s sample size is adequate to detect an effect is an important first step in all analyses, but this procedure is complicated when performing factor analysis. Many sources recommend a general rule of thumb (e.g., five participants per each item in the analysis), and although this approach has been criticized, it is still frequently used within the literature (29, 40). There are 22 items in the 4S, suggesting that a sample size of at least 110 participants is adequate for the full measure, and thus it is safe to use MLE. Univariate and multivariate normality were also tested by observing skewness and kurtosis of individual items and by checking for outliers. Participants’ raw responses were converted into Z-scores, and any participant with more than one response greater than three or less than negative three was dropped. Once this was complete, all items had skewness and kurtosis within the recommended limits of +1.5/-1.5 (47).

72

Figure A1. Preliminary model run on data collected at pre-test, tested independently on pre-test, post-test control, and post-test intervention students.

73

Figure A2. Measurement invariance testing uses an iterative process in which each model is increasingly constrained. The fit statistics are used to evaluate whether the proposed model, with the proposed constraints, fits both groups well. Newly constrained values are added in black, while constraints carrying over from the previous model are in gray; values not depicted in the illustration are free to vary in that iteration of the model.

74

Figure A3. Final model run on data collected at post-test, comparing control and intervention students.

75

Table A1. Comparing factor loadings across three groups (pre-test, post-test control, and post-test intervention) Pre-Test

Post-Test Control

Post-Test Intervention

φ

0.72

0.86

0.81

λ1

0.60

0.63

0.76

λ2

0.59

0.67

0.69

λ3

0.44

0.45

0.58

λ4

0.69

0.75

0.69

λ5

0.59

0.53

0.60

λ6

0.64

0.87

0.71

λ7

0.57

0.78

0.59

λ8

0.62

0.71

0.69

λ9

0.60

0.65

0.78

λ10

0.64

0.78

0.75

λ11

0.61

0.78

0.61

λ12

0.66

0.66

0.70

λ13

0.66

0.77

0.76

λ14

0.60

0.78

0.73

λ15

0.64

0.69

0.67

λ16

0.55

0.75

0.72

76

Table A2. Comparing intercepts across two post-test groups (post-test control and post-test intervention) Post-Test Control

Post-Test Intervention

τ1

-0.13

0.04

τ2

-0.07

-0.01

τ3

0.18

0.08

τ4

0.17

0.13

τ5

-0.08

0.12

τ6

0.01

0.33

τ7

0.05

0.10

τ8

0.11

0.36

τ9

0.11

0.21

τ10

-0.05

0.18

τ11

0.03

0.15

τ12

0.03

0.22

τ13

0.07

0.09

τ14

0.17

0.05

τ15

0.13

0.17

τ16

0.13

0.07

77

Table A3. Comparing variances across three groups (pre-test, post-test control, and post-test intervention) Error Variances

Pre-Test

Post-Test Control

Post-Test Intervention

δ1

0.64

0.61

0.42

δ2

0.65

0.56

0.52

δ3

0.80

0.80

0.68

δ4

0.53

0.43

0.52

δ5

0.66

0.73

0.64

δ6

0.60

0.23

0.49

δ7

0.68

0.39

0.62

δ8

0.62

0.50

0.53

δ9

0.64

0.58

0.36

δ10

0.60

0.39

0.40

δ11

0.63

0.39

0.63

δ12

0.56

0.56

0.52

δ13

0.57

0.39

0.42

δ14

0.64

0.40

0.50

δ15

0.59

0.52

0.55

δ16

0.70

0.43

0.50

Error Co-variances δ12

0.26

0.17

0.24

δ35

0.27

0.43

0.31

δ46

0.22

-0.08

-0.04

δ78

0.15

0.17

-0.06

δ79

-0.16

0.04

-0.34

δ89

0.13

0.06

0.05

δ1013

0.11

-0.06

-0.22

δ1112

0.30

0.48

0.02

δ1415

0.09

0.09

0.16

δ1416

-0.15

0.01

0.07

δ1516

0.08

0.31

0.27

78

References 1.

2. 3.

4.

5.

6.

7.

8.

9.

10. 11.

12.

13.

14.

National Science Foundation; National Center for Science and Engineering Statistics. Women, Minorities, and Persons with Disabilities in Science and Engineering: 2017. Special Report NSF 17-310; 2017. http://www.nsf.gov/ statistics/wmpd/ (accessed May 1, 2018). Gates, S. J.; Mirkin, C. Engage to Excel. Science 2012, 335, 1545. Chen, X. STEM Attrition: College Students’ Path Into and Out of STEM Fields; Statistical Analysis Report NCES 2014-00; NCES: Washington, DC, 2013; pp 35−50. Hill, C.; Corbett, C.; St. Rose, A. Why So Few? Women in Science, Technology, Engineering, and Mathematics; American Association of University Women: Washington, DC, 2010; pp 1−29. Williams, J. C.; Phillips, K. W.; Hall, E. V. Double Jeopardy? Gender Bias against Women of Color in Science; Hastings College of the Law, Center for Worklife Law: 2014; pp 11−48. http://www.uchastings.edu/news/articles/ 2015/01/double-jeopardy-report.pdf (accessed May 1, 2018). Sithole, A.; Chiyaka, E. T.; McCarthy, P.; Mupinga, D. M.; Bucklein, B. K.; Kibirige, J. Student Attraction, Persistence and Retention in STEM Programs: Successes and Continuing Challenges. High. Educ. Stud. 2017, 7, 46. Sowell, R.; Allum, J.; Okahana, H. Doctoral Initiative on Minority Attrition and Completion; Council of Graduate Schools: Washington, DC, 2015; pp 60−66. Fiske, S. T.; Cuddy, A. J. C.; Glick, P.; Xu, J. A Model of (Often Mixed) Stereotype Content: Competence and Warmth Respectively Follow from Perceived Status and Competition. J. Pers. Soc. Psychol. 2002, 82, 878–902. Boysen, G. A. Confronting Math Stereotypes in the Classroom: Its Effect on Female College Students’ Sexism and Perceptions of Confronters. Sex Roles 2013, 69, 297–307. McGee, E. O. Devalued Black and Latino Racial Identities. Am. Educ. Res. J. 2016, 53, 1626–1662. Palmer, R. T.; Maramba, D. C.; Dancy, T. E., II. A Qualitative Investigation of Factors Promoting the Retention and Persistence of Students of Color in STEM. J. Negro Educ. 2011, 80, 491–504. Cheryan, S.; Drury, B. J.; Vichayapai, M. Enduring Influence of Stereotypical Computer Science Role Models on Women’s Academic Aspirations. Psychol. Women Q. 2013, 37, 72–79. Leslie, S.-J.; Cimpian, A.; Meyer, M.; Freeland, E. Expectations of Briliance Underlie Gender Distributions across Academic Disciplines. Science 2015, 347, 23–34. Schinske, J.; Perkins, H.; Snyder, A.; Wyer, M. Scientist Spotlight Homework Assignments Shift Students’ Stereotypes of Scientists and Enhance Science Identity in a Diverse Introductory Science Class. CBE Life Sci. Educ. 2016, 15, p.ar47. 79

15. Andersen, H. M.; Krogh, L. B.; Lykkegaard, E. Identity Matching to Scientists: Differences That Make a Difference? Res. Sci. Educ. 2013, 44, 439–460. 16. Krogh, L. B.; Andersen, H. M. “Actually, I May Be Clever Enough to Do It”. Using Identity as a Lens to Investigate Students’ Trajectories Towards Science and University. Res. Sci. Educ. 2013, 43, 711–731. 17. Oyserman, D.; Destin, M.; Novin, S. The Context-Sensitive Future Self: Possible Selves Motivate in Context, Not Otherwise. Self Identity 2015, 14, 173–188. 18. Diekman, A. B.; Brown, E. R.; Johnston, A. M.; Clark, E. K. Seeking Congruity between Goals and Roles: A New Look at Why Women Opt out of Science, Technology, Engineering, and Mathematics Careers. Psychol. Sci. 2010, 21, 1051–1057. 19. Oyserman, D. Identity-Based Motivation. In Emerging Trends in the Social and Behavioral Sciences; John Wiley & Sons, Inc.: Hoboken, NJ, U.S.A., 2015; pp 1–11. 20. Dasgupta, N. Ingroup Experts and Peers as Social Vaccines Who Inoculate the Self-Concept: The Stereotype Inoculation Model. Psychol. Inq. 2011, 22, 231–246. 21. Stout, J. G.; Dasgupta, N.; Hunsinger, M.; McManus, M. A. STEMing the Tide: Using Ingroup Experts to Inoculate Women’s Self-Concept in Science, Technology, Engineering, and Mathematics (STEM). J. Pers. Soc. Psychol. 2011, 100, 255–270. 22. Corcoran, K.; Mussweiler, T. The Cognitive Miser’s Perspective: Social Comparison as a Heuristic in Self-Judgements. Eur. Rev. Soc. Psychol. 2010, 21, 78–113. 23. Oyserman, D. Identity-Based Motivation: Core Processes and Intervention Examples. In Advances in Motivation and Achievement: Motivational Interventions.; Karabenick, S., Urdan, T. C., Eds.; Emerald Group Publishing Limited: Bradford, U.K., 2014; pp 213–242. 24. Macrae, C. N.; Milne, A. B.; Bodenhausen, G. V. Stereotypes as EnergySaving Devices: A Peek inside the Cognitive Toolbox. J. Pers. Soc. Psychol. 1994, 66, 37–47. 25. Fiske, S. T.; Taylor, S. E. Social Cognition: From Brains to Culture; Sage: Thousand Oaks, CA; 2017. 26. Cheryan, S.; Master, A.; Meltzoff, A. N. Cultural Stereotypes as Gatekeepers: Increasing Girls’ Interest in Computer Science and Engineering by Diversifying Stereotypes. Front. Psychol. 2015, 6, 1–8. 27. Master, A.; Cheryan, S.; Meltzoff, A. N. Computing Whether She Belongs : Stereotypes Undermine Girls’ Interest and Sense of Belonging in Computer Science. J. Educ. Psychol. 2015, 108, 424. 28. Fiske, S. T.; Dupree, C. Gaining Trust as Well as Respect in Communicating to Motivated Audiences about Science Topics. Proc. Natl. Acad. Sci. U.S.A. 2014, 111, 13593–13597. 29. Schinske, J.; Cardenas, M.; Kaliangara, J. Uncovering Scientist Stereotypes and Their Relationships with Student Race and Student Success in a Diverse, Community College Setting. CBE Life Sci. Educ. 2015, 14, 14:ar35. 80

30. Yong, A.-G.; Pearce, S. A Beginner’s Guide to Factor Analysis: Focusing on Exploratory Factor Analysis. Tutor. Quant. Methods Psychol. 2013, 9, 79–94. 31. Brown, T. Confirmatory Factor Analysis for Applied Research, 2nd ed.; The Guilford Press: New York, NY, 2015. 32. Nassar-McMillan, S. C.; Wyer, M.; Oliver-Hoyo, M.; Schneider, J. New Tools for Examining Undergraduate Students’ STEM Stereotypes: Implications for Women and Other Underrepresented Groups. New Dir. Institutional Res. 2011, 2011, 87–98. 33. Wyer, M.; Schneider, J.; Nassar-McMillan, S.; Oliver-Hoyo, M. Capturing Stereotypes: Developing a Scale to Explore U.S. College Students’ Images of Science and Scientists. Int. J. Gender, Sci. Technol. 2010, 2, 382–415. 34. Fox, R. S.; Lillis, T. A.; Gerhart, J.; Hoerger, M.; Duberstein, P. Multiple Group Confirmatory Factor Analysis of the DASS-21 Depression and Anxiety Scales: How Do They Perform in a Cancer Sample? Psychol. Rep. 2017, 121, 548–565. 35. Riegle-Crumb, C.; King, B. Questioning a White Male Advantage in STEM: Examining Disparities in College Major by Gender and Race/Ethnicity. Educ. Res. 2010, 39, 656–664. 36. Shadish, W. R.; Cook, T. D.; Campbell, D. T. Experimental and QuasiExperimental for Generalized Designs Causal Inference; Houghton Mifflin Company: Boston, MA, 2002. 37. Hansen, B.; Fredrickson, M.; Pinelis, Y. Matching in R Using the Optmatch and RItools Packages Optimal Pair Matching and 1 : K Matching; 2013, pp 1−13. https://cran.r-project.org/web/packages/optmatch/. 38. Perkins, H.; Schinske, J.; Wyer, M. Scientist Spotlights: Diverse Images of Scientists. http://scispotlights.hthrperkins.com/ (accessed Mar. 11, 2018). 39. Rosseel, Y. Lavaan: An R Package for Structural Equation Modeling. J. Stat. Softw. 2012, 48, 1–36. 40. semTools Contributors. semTools: Useful Tools for Structural Equation Modeling. Comprehensive R Archive Network (CRAN); 2016; pp 49−50. http://cran.r-project.org/package=semTools (accessed May 1, 2018). 41. Schmitt, T. A. Current Methodological Considerations in Exploratory and Confirmatory Factor Analysis. J. Psychoeduc. Assess. 2011, 29, 304–321. 42. Jackson, D. L.; Gillaspy, J. A.; Purc-Stephenson, R. Reporting Practices in Confirmatory Factor Analysis: An Overview and Some Recommendations. Psychol. Methods 2009, 14, 6–23. 43. Costello, A. B.; Osborne, J. W. Best Practices in Exploratory Factor Analysis: Four Recommendations for Getting the Most From Your Analysis. Pract. Assessment, Res. Educ. 2005, 10, 1–9. 44. Tabachnick, B. G.; Fidell, L. S. Using Multivariate Statistics; Pearson Education: New York, NY, 2013. 45. Subramaniam, K.; Esprívalo Harrell, P.; Wojnowski, D. Analyzing Prospective Teachers’ Images of Scientists Using Positive, Negative and Stereotypical Images of Scientists. Res. Sci. Technol. Educ. 2013, 31, 66–89. 81

46. Miele, B. E. Using the Draw-a-Scientist Test for Inquiry and Evaluation. J. Coll. Sci. Teach. 2014, 43, 36–40. 47. Hashimoto, T.; Karasawa, K. Science, so Close and yet so Far Away: How People View Science, Science Subjects, and Scientists. In Recent Advances in Natural Computing; Suzuki, Y., Hagiya, M., Ed.; Springer Japan: 2015; Vol. 9, pp 57–67. 48. Beardslee, D. C.; O’Dowd, D. D. The College-Student Image of the Scientist. Science 1961, 133, 997–1001. 49. Margolis, J.; Fisher, A.; Miller, F. The Anatomy of Interest : Women in Undergraduate Computer Science. Women’s Stud. Q. 2000, 28, 1–35. 50. Diekman, A. B.; Fuesting, M. A. Choice, Context, and Constraint: When and Why Do Women Disengage From STEM? In APA Handbook of the Psychology of Women; APA: Washington, DC, 2017. 51. McNeill, K. L.; Vaughn, M. H. Urban High School Students’ Critical Science Agency: Conceptual Understandings and Environmental Actions Around Climate Change. Res. Sci. Educ. 2012, 42, 373–399. 52. Godwin, A.; Potvin, G. Fostering Female Belongingness in Engineering through the Lens of Critical Engineering Agency. Int. J. Eng. Educ. 2015, 31, 938–952.

82

Chapter 5

UWM STEM CELL – Accelerating the Pace to Academic Success Anja Blecking* Department of Chemistry and Biochemistry, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin 53201, United States *E-mail: [email protected]

The STEM CELL pilot project program as implemented at the UW-Milwaukee is an innovative project that seeks to transform the first-year experience for aspirational STEM majors. The project, developed in an effort to support STEM-interested students in need of remediation more effectively, and with this build a model that delivers STEM education more effectively during the first-year. This pilot program is meant to provide a template for early career STEM education, focusing on student outcomes. The removal of old existing prerequisite structures and the implementation of an effective students support program has shown to accelerate the pace at which students make academic progress. The STEM CELL program combines several high-impact interventions, such as intentional cross-disciplinary course alignment, in-class student support services, and the construction of individual development plans.

The STEM CELL Program at UWM describes the pilot implementation of a transformational framework to address pedagogical and socio-cultural issues at the University of Wisconsin-Milwaukee (UWM) that contribute to the lack of persistence and the flow of talented and interested students away from science, technology, engineering and mathematics (STEM) majors at the early-career undergraduate level. Students entering UWM with STEM aspirations face different challenges relative to students at more selective institutions.

© 2018 American Chemical Society

Some of these are intrinsic to the university, such as gateway courses taught in the traditional lecture format that fail to provide the intellectual engagement needed to help students successfully persist in completing their STEM majors. Others lie with the student population UWM serves, as almost half of entering students are the first in their family to attend college, and 25% are from an underrepresented racial/ethnic minority group. Many aspirational STEM students entering UW-Milwaukee (i) lack the cultural knowledge about how to operate optimally in a university setting; (ii) have little idea what a STEM career entails; and (iii) have academic backgrounds that are deficient (1). The STEM CELL project has been implemented to fundamentally reorient the institutional culture from an attrition model to one that is focused on student outcomes including STEM persistence; it supports the success of students in need of remediation and provides explicit guidance to students regarding STEM major and career choices; and immerses these students in early academic career experiences that will help them effectively narrow their chosen field of study and career paths. The program intends to establish a model that delivers STEM education in an effective, efficient manner that maximizes persistence, learning, and enthusiasm while minimizing time to degree.

UWM STEM CELL – Program Features Removing Barriers – Changing the STEM Pathway In fall 2015, UW-Milwaukee implemented a pilot project into the undergraduate STEM pipeline to erase a fundamental flaw in the STEM program structure. The program design required students to decide at the end of the first semester if they would like to (i) pursue the chemistry/ engineering/physics track or (ii) pursue the biological sciences track. Depending on their decision, their next math course was either (i) pre-calculus followed by the engineering calculus series or in case students chose the biological science track (ii), they needed to enroll in a survey of calculus course as a terminal mathematics course. In either case, students needed to decide without ever having taken a college science course. The arrow in Figure 1 indicates the student decision point before the implementation of the STEM CELL program. The main reason for this course sequence is the old pre-requisite structure; Intermediate Algebra, Preparatory Chemistry, and Foundations of Biological Sciences I needed to be completed in sequence. For a pilot cohort of 75 first-year students, these interdisciplinary pre-requisites were removed so that students were able to enroll in all three courses concurrently. This new pathway for STEM-intended majors enables students to proceed with their specific program courses earlier and are on a faster track towards graduation (Figure 2). 84

Figure 1. Pre-STEM CELL Program Structure (showing math, chemistry, biological science course sequence, not complete student schedule)

Figure 2. STEM CELL Program Structure (showing math, chemistry, biological science course sequence, not complete student schedule)

The new structure, as shown in Figure 2, moved the decision point so that now students decide after one semester which pathway they would like to pursue. Figure 2 shows the suggested 2-year program pathway of a STEM intended student pursuing a major in either chemistry, biochemistry, or biological sciences). Students choosing the physics or engineering route will enroll in other math and science courses after year 1 to meet their specific program requirements. The program cohort structure spans two semesters. In Semester 1 (fall semester), students are enrolled in Intermediate Algebra, Preparatory Chemistry, and Foundations in Biological Sciences I. In Semester 2 (spring), students take Pre-calculus/Survey of Calculus, General Chemistry I, and Foundations of Biological Sciences II.

85

The more condensed program structure combined with effective academic support and advising is expected to not just save students time to graduation, but also reduce the number of credits and cost to degree completion for these STEM students. Institutional data (UWM) show, e.g. that the average biological science major at UWM accumulates 147 credits, much more than the 120 required credits. How does the STEM CELL program work? How can it support incoming students effectively? The new program structure borrows elements from the highly successful ASAP initiative (Accelerated Study in Associate Programs) at the City University of New York (CUNY) that started in 2007 and has been found to double students graduation rate in comparison to control group students (2, 3). The STEM CELL pilot program has not adopted the ASAP program in its entirety due to the differences in institutions, majors it supports, available services and resources. Instead, it adapted selected key components of CUNY’s student support structure that could be effectively implemented and sustained long-term. Table 1 provides a comparative view of key program features in both programs.

Table 1. Comparison of program core elements; CUNYs ASAP program and STEM CELL program

86

Preparation for Pilot Project Implementation Course Alignment In the summer before the first cohort entered the program, the project team and instructors of all involved fall courses met during a one-week summer workshop to align the course curriculum in ways that seemed most beneficial to students and student learning. Through in-depth discussions, curriculum redundancies were eliminated, concepts aligned, instructional lessons and units rewritten, and schedules finalized. The curriculum in the Preparatory Chemistry course needed to be rearranged to more effectively match the sequence of concepts taught in the biology course. As one example, the concept of intermolecular forces has intentionally been covered in preparatory chemistry before discussing the same concept in the biological sciences course. Additionally, chemistry concepts that require students to solve algebraic equations and work with exponential functions were delayed into the second half of the semester to make sure students acquired these skills beforehand. The concept alignment allowed instructors to explicitly make interdisciplinary connections and choose cross-disciplinary practice examples to highlight broader applications. Midterm and final exam dates were rearranged to avoid scheduling conflicts, such as e.g. multiple exams on the same day. Rescheduling also allowed arranging course specific academic support activities for each of the three courses during the career seminar (exam reviews). The interdisciplinary discussion extended into teaching practices as well, focusing mostly on student-centered teaching strategies (4–6). These practices are long known to foster student learning. While the project team favors these practices, it has been left to each STEM CELL instructor to which extent these strategies are utilized.

Student Selection Eligibility for the STEM CELL cohort depends on the students’ performance results on the UWM chemistry and math entrance exams. STEM-interested students who upon college entry tested into introductory level chemistry and intermediate algebra and need to take the foundational biological science sequence for their major are advised by their college advisor to enroll consecutively in the special sections for all three these courses. Additionally, students are advised to enroll and the supporting career seminar. That said, it is important to mention that students are not required to follow this pathway, just advised. All students have the option to choose alternative pathways.

87

Cohort Model National data show that less than 40% of US students entering the STEM pipeline and just 20% of STEM-interested underrepresented minority students complete a STEM degree (7). The 4-year completion rate for these underrepresented students is low. African American, Latino, and Native American students show completion rates of 13.2, 15.9, and 14 percent while their white and Asian American counterparts complete STEM degrees at much higher rates of 24.5 and 32.4 percent (8). Studies show a variety of reasons for students leaving STEM programs from uninspiring, large lectures, unwelcoming atmospheres from faculty teaching STEM courses, more intense workload in comparison to non-STEM majors to finding themselves insufficiently prepared, especially in math (9, 10). Given these facts, STEM CELL adopted the concept of a cohort-driven learning model combined with strong advising and academic support for the pilot project (11). Long lasting learning communities have been identified to be one of the high-impact practices that positively influence student retention outcomes, especially for traditionally underserved students (12). Learning communities have shown to help “bridge the gap between what students bring to college, and what they expect to take with them when they leave” (13). Students in the STEM CELL learning community are connected through coursework, shared goals, and resources. The community structure provides space for socializing, networking, and academic support (14). Students entering the STEM CELL program are placed in reserved sections of Preparatory Chemistry (lecture and discussion), Algebra, and the Biological Sciences (only lab sections) during the fall semester. During the spring semester, the cohort is placed in the same sections as their non-STEM CELL peers. The vehicle for additional career-focused student support activities is the accompanying STEM CELL one-credit career seminar. Career Seminar The STEM CELL program makes a conscious effort to align resources and processes all first-year students ordinarily have difficulty to maneuver on their own (15–17). Making these resources available and strengthening connections amongst students and between students and faculty is an important part of the program. The career seminar is offered in both semesters of the program once per week. It provides content-specific support and further teaches students critical skills needed to be successful in a STEM program. It is designed and led by STEM faculty and offers support and activities consistent with practices believed to have a positive effect on student retention (18). The seminar is part of the comprehensive student support system teaching students critical skills, providing information about the current STEM job environment, and informing students on how to network with STEM professionals inside and outside the University. This last skill is particularly important for underrepresented students whose social network tends not to be relevant for college and career success (19). 88

Setting the Stage for Collaboration Students in the STEM CELL seminar discuss course expectations with the course instructor during the first week of classes. The starting point for the discussion is a list of instructor and student expectations listed in the career seminar syllabus (Table 2). This discussion is necessary to establish collaborative goals for the semester, and it emphasizes that the focus of the seminar is student learning and student success. It further underscores that students need to have an active part in this process. They have to take responsibility for their learning, exercise self-assessment and accept the support the course offers.

Table 2. Instructor and Student Expectations as noted in STEM CELL seminar syllabus

The goal of the open discussion is for students to recognize that all course activities as outlined in the course schedule are directly aligned with these expectations and goals.

89

Research shows that two of the main factors influencing student motivation are the value of an activity or course for a student and the nature of the course environment (12). Students are less likely to be motivated if they feel unsupported. Optimal learning requires a partnership between teachers and learners, with the expectations of both sides clearly stated. Unclear goals and expectations can lead to frustration, miscommunication, and a decrease of motivation for teachers and students (20).

Explicit Career Advising and Construction of Individual Development Plans (IDP) Individual Development Plans (IDPs) are a tool utilized in a variety of professional settings, e.g., teacher education programs, undergraduate and graduate institutions, professional development in industry and government agencies. Gollwitzer (21) stated that “by forming goal intentions, people translate their noncommittal desires into binding goals” and are more likely to follow up with action steps necessary to reach these goals. He further explains that the path to attain one’s goals depends on the determination of behaviors or skills needed. Students in the STEM CELL cohort are asked to work on their individual development plan, identifying their short-term (+5 years), intermediate (+10 years), and long-term goals (+15 years). For each of the 5-year segments, students establish a timeline indicating when they plan to achieve these goals and identify the skills needed that will help them on their journey. Lastly, this exercise asked them to identify strategies on how to acquire the important, needed skills. STEM CELL students create IDPs at the beginning of the fall semester with a chance to revise it later in the semester. This exercise forces students to self-reflect on their personal and academic progress to date and to assess what next steps are. Instructors emphasize that when setting individual goals, students need to make sure that timelines are attainable, and goals are specific, measurable, and within their control. Students should feel empowered and eager to seek resources to develop their skill set proactively (22). Individual Development Plans provide a solid framework for conversations with mentors, course instructors, and student advisors. It prepares students to choose opportunities along the way wisely, take advantage of the ones that support their goals and decline other that may delay reaching them. Not surprisingly, even students who enter college with a solid vision of their future after college, know very little about the specific skills they need to reach their goals and even less about the coursework or other resources that will help them acquire these skills. The STEM CELL seminar invites advisors specialized in natural sciences, pre-health and engineering majors regularly to the classroom for in-depth careerfocused discussions The project team found this especially important because STEM-intended students entering UWM face an abundance of choices as to how to schedule their classes, engage in advising, and document their progress. A growing body of research indicates that such a “paradox of choice” frustrates many (16, 23) and structuring choice early in college is likely 90

to significantly improve outcomes (17). Needed student services such as career-focused advising often remain largely separated from classroom learning, remedial coursework is performed outside of the work of disciplines, and the introductory course requirements across multiple departments create an environment of disconnect that works against student success. The career seminar curriculum also includes the discussion of a variety of widely applicable skills that are needed to be successful in college, such as notetaking skills, study skills, time and stress management skills. Throughout the course, students are repeatedly asked to identify and reflect on their skills and learning strategies and share and evaluate them in small groups. Saundra McGuire describes these metacognitive strategies as “overarching principles that enable students to stop failing their classes and start acing them” (20). She further explains that students who employ metacognition become consciously aware of themselves as problem solvers, which enables them to actively seek solutions to any problems they may encounter, rather than relying on others to tell them what to do or to answer their questions (20). The STEM CELL seminar adopted material into for the course curriculum that has shown to foster metacognitive strategies. Concepts such as Carol Dweck’s Growth Mindset (24), the Study Cycle, Bloom’s Taxonomy (25), and post-exam reflection continuously asks students to reflect on their behavior, strategies, and skills. Very important strategies are being repeatedly discussed throughout the semester. The study cycle, e.g. (Table 3) is a frequent discussion topic before exams to provide a guiding model for more effective learning and exam preparation. The 5-step approach also encourages students to use time wisely and be in charge of their learning.

Table 3. Study Cycle Strategies. Adapted with permission from reference (26). Copyright 2015 Louisiana State University. 1. Preview

Preview before lass – Skim the chapter, note heading and boldface words, review summaries and chapter objectives, and note questions you would like answered in class.

2. Action

Be an active listener – ask questions and take meaningful notes.

3. Review

Review after class – within 24 hours, read notes, fill gaps and note any questions.

4. Study

Study – repetition is the key. Ask questions such as “why”, “how”, and “what if”. Intense Study Sessions – 3 to 5 short study sessions per day. Weekend Review – Read notes and material from the week to make connections.

5. Assess

Assess your learning – periodically perform reality checks. Am I using study methods that are effective? Do I understand the material enough to teach it to others?

91

Content Specific Academic Support Exam review sessions for all three courses in the fall and spring semester of the STEM CELL cohort are conducted by the chemistry, math, or biology course instructors as part of the seminar course. To maximize student learning, reviews employ active-learning strategies. Students often work on problem-solving strategies in small groups, then share and discuss results with the larger class.

Exploration of STEM Careers As previously mentioned, one of the STEM CELL seminar’s foci is to provide advising and career-relevant information. To give students more insight into STEM careers and the skills and practices needed to be successful in these careers, the seminar invites guest speakers in STEM careers relevant to the student’s career plans. Students are being encouraged to interact with these professionals and ask questions. Their work on their own Individual Development Plan can provide a helpful framework for these discussions.

Early Warning System STEM cell course instructors monitor student progress closely. They meet weekly to share student assessment data, discuss individual student growth, and discuss support strategies for struggling students. Instructors meet with students who receive exam or homework grades below 70% to discuss ways on how to improve. The weekly “project check-ins” are also critical to ensure that the course content alignment throughout the semester is on target.

Student Feedback The STEM CELL project team collects student feedback on seminar activities regularly to evaluate how these are valued. An end–of semester survey investigates the main seminar topics and activities mentioned below. Data show that the majority of students rated the value of these activities very positively. Seminar activities rated: • • • • • • •

Review Sessions Study Skills Notetaking Stress Reduction Reading Skills Time Management Seeking Resources 92

• •

Individual Development Plan (IDP) Invited Speakers

While a large number of students found most activities beneficial, others did not see any value discussing skills, such as note-taking, study skills, or time management, which in their opinion were skills they acquired before setting foot on a college campus. Although the project team is very receptive to student feedback, it was decided to keep these components in the career seminar curriculum, rating research findings higher than student opinion.

Additional Student Comments about Seminar Activities Individual Development Plan • “It was a good activity that made me think about what I really want to do in the future; it allowed me to see what I need to do in order to achieve my goals.” • “Keep the discussions about short-term academic future and which classes to take in the following semesters.” • “The template was useful.” • “I found it to an awkward assignment at first. Like many of the college students I know, I have no idea what I am even doing tomorrow, much less ten years from now, but it was very good to start thinking about it.” Invited Speakers • “Speakers provided helpful tips for our future.” • “The speakers were very passionate about science.” • “Speakers who talked about their academic path in the science majors, med school, and the time and effort it takes to be a resident at a hospital were very helpful.” Course Comments • “…the course helped me a lot and provided guidance throughout my first semester.” • “The course was very interesting, and the concept really paid off. “

Results, Project Sustainability, and Scalability This pilot study provided preliminary results that showed the STEM CELL program facilitates students progressing through their introductory courses in less time compared to prior-year cohorts. The change in the program structure and the removal of the existing prerequisite structure along with the newly implemented support allow a larger percentage of students to advance at a faster rate and potentially complete a STEM degree in less time. In addition, working with advisors both created advocacy and support for the program as well as support to 93

the students keeping them on their academic path. Finally, the cohort model and first-year seminar course provided academic and learning support to increase the probability of student success. Based on the results the STEM CELL pilot project, the existing prerequisite structure will be removed for all sections of the courses allowing all students to take Introductory Chemistry, Intermediate Algebra, and Foundations of Biological Sciences concurrently. Furthermore, the success of the STEM CELL model might encourage wider adoption of this approach into other disciplines, or institutions.

Acknowledgments The author would like to thank the entire STEM CELL project team, past and present, especially Dr. Kyle Swanson (Metropolitan State University, St. Paul, MN, formerly UWM), Dr. Peter Geissinger (Eastern Oregon University, La Grande, OR, formerly UWM), Dr. Kristen Murphy and Dr. Daad Saffarini (both at UW, Milwaukee, WI), as well as STEM CELL course instructors, tutors, advisors, and participating UWM students.

References 1. 2.

3. 4. 5.

6.

7. 8.

UW-Milwaukee, Institutional Data, 2015. Scrivener, S.; Weiss, M. J.; Ratledge, A.; Rudd, T.; Sommo, C.; Fresques, H. Doubling Graduation Rates: Three Year Effects of CUNY’s Accelerated Study in Associated Programs (ASAP) for Developmental Education Students; MSRC: New York, NY, 2015. Retrieved from: http://www.mdrc.org/sites/default/files/doubling_graduation_ratesfr.pdf. Scrivener, S.; Weiss, M. J.; Sommo, C.; What can a multifaceted program do for community college students? MDRC: New York, NY, 2012. Seymour, E.; Hewitt, N. M. Talking about leaving: Why undergraduates leave the sciences; Westview Press: Boulder, CO, 1997. Goodman Research Group Inc. (Goodman I. F.; Cunningham, C. M.; Lachapelle, C.; Thompson, M.; Bittinger, K.; Brennan, R. T.; Delci, M.; Final Report of the Women’s Experiences in College Engineering (WECE) Project; Goodman Research Group: Cambridge, MA, 2002. Watkins, J.; Mazur, E. Retaining students in science, technology, engineering, and mathematics (STEM) majors. J. Coll. Sci. Teach. 2013, 42, 36–41. President’s Council of Advisors on Science and Technology (PCAST). Engage to Excel; White House Press: Washington, DC, 2012. Higher Education Research Institute (HERI). Degrees of success: Bachelors degree completion rates among initial STEM majors, 2010. Retrieved from: http://www.heri.ucla.edu/nih/downloads/ 2010%20-%20Hurtado,%20Eagan,%20Chang%20-%Degrees%of%20 Success.pdf. 94

9.

10.

11.

12.

13.

14. 15. 16. 17.

18. 19.

20.

21. 22.

23. 24. 25.

26.

Rogers, W. D.; Ford, R. Factors that affect student attitude toward biology. Bioscene 1997, 23, 3–5. Retrieved from http://acube.indstate.edu/ volume_23/v23-2p3-5.pdf. Cheryan, S.; Master, A.; Meltzoff, A. N. Cultural stereotypes as gatekeepers: Increasing girls’ interest in computer science and engineering by diversifying stereotypes. Front. Psychol. 2015, 6DOI: 10.3389/fpsyg.2015.00049. Scrivener, S.; Weiss, M. J. More graduates: Two-year results from an evaluation of Accelerated Study in Associate Programs (ASAP) for developmental education students; MDRC: New York, NY, 2013. Ambrose, S. A.; Bridges, M. W.; DiPietro, M.; Lovett, M. C.; Norman, M. K. How learning works: Seven research-based principles for smart teaching; Jossey-Bass: San Francisco, CA, 2010. Venezia, A.; Kirst, M. W.; Antonio, A. L. Betraying the College Dream: How disconnected K-12 and postsecondary education systems undermine student aspirations; The Bridge Project: Stanford, CA, 2003. Laufgraben, J. L.; Shapiro, N. S. Sustaining and Improving Learning Communities; Jossey-Bass, Wiley: 2004. Schwartz, B. The paradox of choice: Why more is less; Harper Collins: New York, NY, 2004. Thaler, R. H.; Sunstein, C. R. Improving decisions about health, wealth, and happiness; Yale University Press: New Haven, CT, 2008; p 293. Scott-Clayton, J. The shapeless river: Does a lack of structure inhibit students’ progress at community colleges? Community College Research Center Working Paper No. 25, New York, NY, 2011. Levitz, R. S.; Noel, L.; Richter, B. J. Strategic Moves for Retention Success. New Directions Higher Educ. 1999, 31–49; DOI: 10.1002/he.10803. Rios-Aguilar, C.; Deil-Amen, R. Beyond getting in and fitting in: Leveraging a trajectory of educational success through professionally-relevant social networks. J. Hispanic Higher Educ. 2012, 11, 179–196. McGuire, S. Y. Teach students how to learn: strategies you can incorporate into any course to improve student metacognition, study skills, and motivation; Stylus Publishing: Sterling, VA, 2015. Gollwitzer, P. M. Implementation Intentions-Strong Effects of Simple Plans. Am. Psychol. 1999, 54, 493–503. Kuh, G. D.; O’Donnell, K. Ensuring Quality and Taking High-Impact Practices to Scale; Association of American Colleges and Universities: Washington, DC, 2013. Schwartz, B. The paradox of choice: Why more is less; Harper Collins: New York, NY, 2004. Dweck, C. S. Mindset: The new psychology of success; Ballantine Books: New York, NY, 2016. Bloom, B. S.; Englehart, M. B.; Furst, E. J.; Hill, W. H.; Krathwohl, D. R. Taxonomy of educational objectives: The classification of educational goals; McKay: New York, NY, 1956; p 1. Lousiana State University, Center for Academic Success. The Study Cycle, 2015. Retrieved from http://www.lsu.edu/students/cas/makebettergrades/ note-based.php#Study%20Cycle. 95

Chapter 6

Effective Strategies To Improve Academic Success and Retention in Underrepresented STEM Students Pamela K. Kerrigan,* Ana C. Ribeiro, and Patricia A. Grove Division of Natural Sciences, College of Mount Saint Vincent, Riverdale, New York 10471, United States *E-mail: [email protected]

The College of Mount Saint Vincent (a primarily first generation and/or minority student population) was awarded an NSF S-STEM grant in 2013 to provide financial, social and academic support to improve academic performance, retention, degree completion and post-baccalaureate opportunities for students in biology, chemistry, and biochemistry. A freshman course “Survival Skills for Science” was developed to increase the knowledge base, discipline, and critical thinking skills of students, provide peer mentorship and tutoring for the first 2 years of baccalaureate study; and encourage participation in scientific conferences, career panels, and field excursions to museums and academic centers. A faculty member and a peer leader instructed this freshman seminar. The peer leaders were successful upperclassman who were trained in how to facilitate their sections. The role of the peer leader was to model positive behavior, represent the College with pride and respect, serve as an advocate for the interests, needs, and concerns of first-year CMSV students, help students adjust to college life and become academically successful. The freshman seminar consisted of several modules that included topics such as scientific terminology, graphing, time management, working in groups, writing a laboratory report, researching and presenting a specific topic, among others. The program has confirmed that cohort building, mentorship, and increased academic support can bolster student success. The most significant lesson learned

© 2018 American Chemical Society

was that there is a strong need for continued, intensive academic support in science courses beyond Year 1, when the attrition rate increased due to unsatisfactory performance in science courses.

Introduction The National Science Foundation (NSF) awarded an S-STEM grant to the College of Mount Saint Vincent (CMSV) in 2013 to establish a Scholars-On-Track Program to provide financial, social and academic support, to improve academic performance, retention, degree completion and post-baccalaureate opportunities for students in biology, chemistry, and biochemistry. The student population that declared a STEM major in 2016 at CMSV was 6% first generation and/or 64% minority. CMSV enrolls a large number of low income and first generation college students. In the Fall of 2013, 41.45% of the students’ college wide were first generation and 4.92% of those students majored in the sciences (Table 1). As can be seen in Table 2, in this same year 58.3% of the students who declared biology/biochemistry/chemistry as a major were either self-identified as black or Hispanic.

Table 1. First-Generation Students by Entry Cohort College Wide (% of total cohort)

Biology Major (% of total cohort)

Biochemistry Major (% of total cohort)

Chemistry Major (% of total cohort)

Fall 2011

45.79%

13.90%

0.91%

0.00%

Fall 2012

44.24%

16.67%

0.62%

0.62%

Fall 2013

41.45%

4.40%

0.26%

0.26%

Fall 2014

46.62%

7.74%

0.58%

0.58%

Fall 2015

37.07%

5.61%

0.31%

0.62%

Fall 2016

28.21%

5.05%

0.00%

0.23%

98

Table 2. Students Declaring Biology/Biochemistry/Chemistry Majors: Minority Fraction 2011-16 Entry Cohort

# students admitted

%black or Hispanic

2011

62

66.13%

2012

87

57.47%

2013

60

58.33%

2014

94

59.57%

2015

54

66.67%

2016

77

63.64%

It is not uncommon for students to change their majors during their academic careers for a variety of reasons. The faculty at CMSV were concerned that a significant determining factor in students changing their majors from a STEM field were due to extreme financial needs (necessitating outside employment), social and family pressures, along with the fact that for some students English is not their first language (1). These disadvantages can lead to students not persisting in the rigorous STEM majors, such as biology, biochemistry or chemistry since these majors require significant time and energy to be successful. The Scholars-on-Track program in the Division of Natural Sciences awarded scholarships based on a combination of merit-based and need-based financial assistance. The program complemented other successful programs already offered at the College, such as our academic support center, but added components that were specifically designed for science students. The program provided the following (a) financial aid so that students would not need to work as many hours at outside employment; (b) activities and services to build necessary skills to be successful as a science major, such as a freshman seminar course; (c) a community building workplace (scholars office); (d) summer research experience and attendance at national and regional conferences and (e) career preparation in the form of internships and mentorship. All CMSV’s first-semester freshmen take the one-credit First Year Experience (FYE) 101, a freshman orientation seminar, to ease their transition into college. However, this course does not address the specific needs of science majors. Science faculty at CMSV report that students are generally poorly prepared for the rigors of college-level science courses (2). Their verbal and quantitative skills are challenged by the course texts and assignments. Regrettably, most students were unchallenged by secondary school, and as a consequence, have not developed the attitude of engagement and study habits necessary for success in college-level science courses. The aim of this study was to determine if financial assistance and increased academic, research and mentorship support would increase retention rates of underrepresented minorities in STEM majors.

99

Methods Our program created a freshman seminar course entitled “Survival Skills for Science” whose goals were to develop the knowledge base, discipline, and critical thinking skills of students; and to enhance their performance in introductory science courses by providing peer mentorship and tutoring (3). A faculty member and a peer leader led these seminar classes. This course involved the creation of several modules that addressed topics such as: scientific terminology (vocabulary-building), active learning in class, note taking, time management, effective study group strategies, textbook reading skills, test-taking strategies (reading skills and critical thinking skills), improving performance in laboratory experiences, proper writing of laboratory reports, graphing, quantitative skills, effective oral and written communication. This course was not meant to redress all shortfalls in preparation for college science, rather it had the goal of helping students develop habits that would enhance their learning, confidence, and performance in their current and future science courses. The two assignments used to improve their writing skills were to write a formal laboratory report and a group research paper. For the laboratory report they were to pick an experiment from either their General Biology or General Chemistry course in which quantitative data was collected and had to write the formal report. This report would include an introduction, methods, results, discussion and reference sections, following the format that is used in freshman and sophomore laboratory courses in the Division of Natural Sciences. They were required to have at least 3-4 pages, not including graphs, tables or figures. During the semester these students were required to do this same assignment in their General Biology or Chemistry course, so it gave them an opportunity to have additional feedback and time to work on their final report for class. The research project had two components: a paper and an oral presentation. Students worked in groups of 2 or 3. The project was based on broad topics, for instance: “What happens to plastics when they are thrown them away?” Each member of the group picked a subtopic related to the general theme such as “Biodegradable plastic, recycling of plastics, environmental effects of plastics” to research. All research projects had to be approved by the faculty mentor. This exercised helped students work on their individual sections and also develop their interpersonal skills working as a group and presenting their work to the class. Each student of the group wrote 2-3 pages on their sub-topic ad the group then had to meet to combine their information to make a cohesive paper illustrating the connections of all sub-topics. Once the sub-sections were complete, the group again had to work together to write an introduction and conclusion along with a reference page. In order to improve their communication skills each group was then responsible for preparing a fifteen-minute presentation on their research project. Each individual was responsible for five minutes of the presentation. The theory behind this presentation was to expose them to the type of presentations they will be doing in their upper level courses. In their senior year these students are required to write a 15-20-page literature review paper and deliver a 15-minute oral presentation to the Division faculty and students. 100

One community building opportunity given to the scholars was an office dedicated for their use. It was equipped with a PC computer, a color printer, a desk and a library of science textbooks. All students at the college are given an allotment of printing copies that they use each semester. If they exceed that number they must pay for any additional copies out of pocket. This puts an added burden on science students, who often have more materials to print and review. Scholars did not have to worry about this since they had a printer to use free of charge in their office. This office also became a quiet study area for them and/or a place for a group to have a study session. This office was used on a regular basis by all of the scholars. These students were also provided with a mentor from outside the college during their junior and senior year through the College’s Mentor Program. All Scholars were eligible to participate in summer research with Division faculty between sophomore/junior year and junior/senior year. If they participate in research the grant provided a stipend for the summer and funding to attend a national or regional scientific meeting to present, their research findings. The criteria for choosing peer leaders were that the students were either a sophomore, junior or senior status, and had to have an overall major GPA of at least 3.0. The role of the peer leader was to model positive behavior, serve as an advocate for the interests, needs, and concerns of first-year CMSV students, help students adjust and become academically successful and serve as a role model for academic, co-curricular, and leadership involvement. Interested and qualified students submitted an application, including a brief essay explaining both what they would bring to the position and what they hope to gain. The applications were reviewed and promising candidates were interviewed by the faculty involved in the courses. In addition to assessing the academic abilities the applicants, each candidate’s interpersonal and communications skills was examined. Once selected, peer leaders were trained in group dynamics and effective discussion management. All peer leaders worked directly under the supervision of a faculty member as a teaching assistant. All peer leaders were required to submit a weekly journal regarding their experiences. Upper-level scholars often became peer leaders.

Results The program exceeded many of its stated goals. By administering scholarships on a tiered basis, the program increased the size of the proposed cohort. First-year enrollment was 14 students, in excess of the 10 in the objective, with another 9 students enrolled in Year 2 and 3 students enrolled in Year 3, for a total of 26 participants, above the goal of 22. Ten of the 14 students enrolled in Year 1 are expected to graduate in 2017 for a retention rate of at 71.4%, which is a significant increase above the baseline rate of 36.8%, but one which needs additional support to reach the revised program goal of 90%. Sixty eight percent of participants achieved a B or higher in their first-year science courses, falling short of the revised goal of 90%. 101

Fifty-four percent completed the first two years with an average of B or better, highlighting the need for increased academic support after Year 1 to meet the revised goal of 90%. It should be noted, however, that the program did increase success rates to well above the TRIO rate of 50%, an objective the grant reviewers encouraged the college to aim for. The program also showed significantly higher letter grades and over GPAs in science courses for participants compared with other students in first and second-year science courses (Tables 3, 4 and 5) Overall, the number of STEM students at CMSV transition from Year 1 to Year 2 decreased by approximately 60% (Table 6) illustrating the need for interventions at this critical time. Table 5 illustrates that the total number of students declaring a STEM major decreases from 78 to 44 for the Class of 2016 and from 54 to 33 from 2017. This is not simply a STEM problem, as evidenced by the increased efforts to overcome the “sophomore slump” (4). The program also showed significantly higher GPAs in science courses for Scholars, compared with other students in first- and second-year science courses.

Table 3. S-STEM Cohort 1 Grades in Core Science Curriculum. Scholars GPA(Scholars on Track scholarship recipients), Class GPA (non-scholarship students); Scholars GPA minus Class GPAa Cohort 1

Scholars GPA

Class GPA

Difference Scholars vs Class

BIOL 111

3.14

1.84

1.30

BIOL 112

3.18

2.42

0.76

CHEM 120

2.92

1.91

1.01

CHEM 121

2.93

1.05

1.88

BIOL 217

3.54

2.64

0.90

BIOL 223

3.27

1.2

2.07

CHEM 219

3.53

2.05

1.48

CHEM 220

3.30

1.78

1.52

PHYS 207

3.75

2.31

1.44

PHYS 208

3.54

2.38

1.16

F13/S14

F14/S15

F15/S16

a

BIOL 111/112 – General Biology I and II; CHEM 120/121 General Chemistry I & II; BIOL 217 Genetics; BIOL 223 Ecology; CHEM 219/220 Organic Chemistry I & II, PHYS 207/208 General Physics I & II.

102

Table 4. S-STEM Cohort 2 Grades in Core Science Curriculum. Scholars GPA(Scholars on Track scholarship recipients), Class GPA (non-scholarship students); Scholars GPA minus Class GPAa Cohort 2

Scholars GPA

Class GPA

Difference Scholars vs class

BIOL 111

3.03

1.28

1.75

BIOL 112

2.50

2.07

0.43

CHEM 120

2.99

1.63

1.36

CHEM 121

3.04

2.41

0.63

BIOL 217

3.67

2.55

1.12

BIOL 223

3.33

2.83

0.50

CHEM 219

2.87

2.14

0.73

CHEM 220

2.59

2.00

0.59

PHYS 207

4.0

2.82

1.18

PHYS 208

3.67

2.71

0.96

F14/S15

F15/S16

F16/S17

a

BIOL 111/112 – General Biology I and II; CHEM 120/121 General Chemistry I & II; BIOL 217 Genetics; BIOL 223 Ecology; CHEM 219/220 Organic Chemistry I & II, PHYS 207/208 General Physics I & II.

The strategies developed in this study greatly enhanced student retention in STEM fields (71% remained in a STEM field from year 1 to year 4), compared to non-grant participants (16 and 15% of students remained STEM majors from year 1 to year 4 in the 2016 and 2017 graduating class respectively (Table 7). One of the ten students who graduated in 2017 did so as a biology major but was no longer part of the Scholars Program but was included in the 10 students who graduated that year. An unexpected finding of this study was the positive impact that being a peer leader had on student engagement and performance. The peer leaders benefited from the extra review they provided Freshman, and this role helped them develop creative strategies to improve the performance of the students in their course. Many of their journal reflections were very positive and demonstrated their personal satisfaction in helping underclassmen develop their study skills and improve their overall performance.

103

Table 5. Representative Grade Distribution in Mandatory Core Science Courses Over a Three-Year Period.a Scholars

Non-Scholars

A

B

C

D

F

A

B

C

D

F

BIOL 111

7

6

0

0

0

25

54

30

2

4

BIOL 112

3

7

2

0

0

19

47

29

2

2

BIOL 217

7

3

0

0

0

12

34

33

8

2

BIOL 223

5

2

0

0

0

22

35

28

4

1

CHEM 101

6

5

1

0

0

18

46

39

6

6

CHEM 102

6

5

1

0

0

14

44

35

6

4

CHEM 219

8

3

0

0

0

17

35

24

4

6

CHEM 220

6

4

0

0

0

8

24

26

1

2

PHYS 107

7

2

0

0

0

11

35

22

2

2

PHYS 108

6

3

0

0

0

10

23

12

2

1

a BIOL 111/112 – General Biology I and II; CHEM 120/121 General Chemistry I & II; BIOL 217 Genetics; BIOL 223 Ecology; CHEM 219/220 Organic Chemistry I & II, PHYS 207/208 General Physics I & II.

Table 6. All STEM Majors (including scholars)a

a

GPA 3.0 or higher at end of first year (N)

Percentage

GPA 3.0 or higher at end of second year (N)

Percentage

Class of 2016

29 (78)

37.2

19 (44)

43.2

Class of 2017

17 (54)

31.5

15 (33)

45.5

Class of 2018

24 (80)

30.0

n/a

n/a

N = number of declared science majors

104

Table 7. Science Students Who Maintained Major to Graduation Number of Students Who Declared Major in Biology, Chemistry, or Biochemistry at Matriculation

Number of Those Students Who Graduated with a Major in Biology, Chemistry, or Biochemistry

Class of 2016

86

14 (16%)

Class of 2017

60

9 (15%)

Scholars in Class of 2017

14

10 (71%)

The following are some examples of peer leader comments from their weekly journals: “Yesterday during the SI section of FYE I split the class into 4 groups of three and one group of 2. I allowed the students in each group to talk to each other about what things they have in common, why they think science is hard, and something about themselves. Some people got stumbled along the way and became quiet after a few minutes, so I asked them some questions like what was their favorite hobby and I also asked why they think science is hard. Some gave me simple an answer such as science contains a lot of material but I told them that they have to think harder of why science is hard. After helping some of the students they all introduced their classmates to the class.” “I think this week’s Freshmen Seminar class went really well! I noticed the students were much more engaged in the lesson and working with each other really well. They were doing the math/statistics packet in groups of 4, and it seemed like they were getting more comfortable asking each other, myself, and Dr. X for help. I was so happy to see this! They all seemed terrified to speak during the last few classes but now they are starting to come out of their shells. I was in their place not too long ago, so I completely understand why they weren’t as chatty. (When you leave the comfort of your home, town, and high school and you’re thrown into college classes with total strangers it can be SCARY!)” For example, the science prefixes. I am also appreciating the extra math review. I am really enjoying this position as a peer leader! I love being an extra source of guidance and insight for the new students. When I transferred here I felt a little lost and I wish I had a class like this and a peer leader that I could talk to. I really hope they take full advantage of this opportunity! I am also going to start emailing them, as you suggested. Hopefully, they’ll feel more comfortable talking to me!” “I tried my best to stress the importance of this course, and how much I wish I had a class like this when I started out as a freshman. That being said, I think having peer leaders is a GREAT idea. The students definitely feel more comfortable talking to fellow students and I also think they are 105

more apt to take advice from peers rather than professors- simply because we were in their same position not too long ago. I really enjoyed my position as a peer leader. It was rewarding to know that I had a positive and helpful impact on these students.”

Conclusion One objective obtained was to increase the number of students who successfully complete the first two years of a baccalaureate program by 20%. Fifteen of the twenty-six participants completed the first two years of their program, for a rate of 58%, far above the original goal of 31.8%. The overall number and percentage of recipients leaving the program was 50% (n=13). Two students transferred out of the college, four students changed their major, and two students graduated with science degrees but were no longer eligible for the scholars grant. These two students and the remaining five lost their eligibility due to poor science grades. All peer leaders chosen from the first cohort were successful in completing their science degrees with GPA’s greater than 3.0 (Table 8). The scholars in the first cohort were successful in gaining admission to professional schools such as a graduate physician’s assistant program (n=1), dental school (n=1), medical school (n=1), and graduate biochemistry programs, (n=2.). The scholars who graduated in 2017 received all of the division medals in biology, biochemistry and research, as well as graduated with honors.

Table 8. GPA Scholars Scholars

GPA 3.0 or higher at end of first year

Percentagea

GPA 3.0 or higher at end of second year

Percentage

Fall 2013 cohort ’17 (14 entered, 10 continuing)

9 (9)

64.3 (90.0)

10

100

Fall 2014 cohort ’18 (9 entered, 7 continuing)

6 (6)

66.7 (85.7)

n/a

n/a

a Numbers in parentheses refer to Scholars who are continuing in the program to Fall 2015.

It should be emphasized that not only being in the program but also being a peer leader built a sense of “community” and this significantly affected their success. Another successful activity was the trips to science museum, Human Body Exhibit, and attendance at regional and national scientific meetings (American Chemical Society, Sigma Xi). The most significant lesson learned was that there is a strong need for continued and intensive academic support in science courses Year 1, which is 106

when the attrition rate is increased due to insufficient grades (5). In the future, the goal is to reduce attrition by focusing our efforts on continued faculty and peer support as well as enhanced and expanded academic support. The program confirmed that student success can be bolstered by reduced financial burdens on students and increased cohort, mentor, and academic supports.

References 1. 2. 3.

4.

5.

Devall, E. Strategies for Recruiting and Retaining Hispanic Students. Fam. Consumer Sci. Res. J. 2005, 97, 50. Barlow, A.; Villarejo, M. Making a Difference for Minorities: Evaluation of an Educational Enrichment Program. J. Res. Sci. Teach. 2004, 41, 861–881. Rodger, S.; Tremblay, P. The Effects of a Peer Mentoring Program on Academic Success Among First Year University Students. Canadian J. Higher Educ. 2003, 33, 1–18. Wang, X.; Kennedy-Phillips, L. Focusing on the Sophomores: Characteristics Associated With the Academic and Social Involvement of Second-Year College Students. J. Coll. Stud. Dev. 2003, 53, 541–548. Graunke, S.; Woosley, S. An Exploration of the Factors that Affect the Academic Success of College Sophomores. Coll. Stud. J. 2005, 39, 367–376.

107

Chapter 7

Seeking To Improve Retention through Teaching Strategies and Peer Tutoring Tara L. Kishbaugh,*,1 Steve Cessna,1 Lori Leaman,2 and Daniel Showalter3 1Biology

and Chemistry Departments, Eastern Mennonite University, 1200 Park Road, Harrisonburg, Virginia 22802, United States 2Teacher Education Department, Eastern Mennonite University, 1200 Park Road, Harrisonburg, Virginia 22802, United States 3Mathematics Department, Eastern Mennonite University, 1200 Park Road, Harrisonburg, Virginia 22802, United States *E-mail: [email protected]

Student retention within STEM majors is often lower than for other majors and can be particularly low for populations that are traditionally under-represented in STEM fields. The under-represented groups at our institution include women in some programs, first-generation college students, as well as racial and ethnic populations, particularly African-Americans, and Latinos. While our student body has grown more diverse over the past 10 years and include significant numbers of first-generation students, our faculty don’t reflect that diversity. The STEM departments use ‘high-impact’, evidence-based approaches to engage students in science, such as required research by all students, embedded research projects in traditional coursework, active-based learning pedagogies in the classroom and more. Yet, our retention of under-presented minorities (URM) and first-generation (FG) students is significantly worse than for majority students. Thus, we sought to reverse this attrition of students through institutional changes in peer tutoring and faculty development. The STEM peer tutoring program was refocused through increased training, reflective activities, and better connections between tutors and faculty. Additionally, a cohort of STEM and first year writing faculty was provided with a year-long professional learning © 2018 American Chemical Society

community devoted to diversity responsive teaching to increase faculty awareness, appreciation, and use of evidence-based and culturally-responsive teaching strategies. Preliminary findings about these changes point to the complexity of identity development in both students and faculty as well the importance of learning communities for professional growth.

The Problem: Low Retention of URM Students Despite ‘High-Impact Practices’ Eastern Mennonite University is a small liberal arts university in Harrisonburg, VA. While most of EMU’s students (67%) are undergraduates, the school has graduate programs, including a Masters of Arts in Biomedicine. These biomedicine masters students often take cross-listed courses with our biology and chemistry undergraduates. In 2015, there were 1224 undergraduate students enrolled and 945 would be considered traditional undergraduate students. Our STEM majors are a reasonably large portion of this undergraduate student population; for example in 2015, 19% the traditional undergraduate students or 187 students had declared STEM majors. We offer major programs of study in Biology, Biochemistry, Chemistry, Clinical Lab Sciences, Computer Science, Engineering, Environmental Sustainability- Science, and Mathematics. The Engineering program began in fall 2016. The largest STEM major is Biology (116 students, fall of 2016), followed by Computer Science (29), which is not that different from Math, Engineering, and Environmental Science (~20 each). Figure 1 shows how the diversity of our STEM classes has steadily increased over the past 10 years. This shift reflects the diversifying demographics seen across the United States.

Figure 1. Percent of First-Year STEM Students who are URM, 2005-2015 110

This diversification effect seems to be generalized over all of our STEM programs (Figure 2). For example, 50% of the STEM students were either URM or first generation in the fall of 2016 at EMU. About 30% of STEM majors were URM, as defined by racial and ethnic categories. At EMU, most of these students would identify as African, African-American, and/or Latino/a. This is true when STEM majors are separated by departments: Biology, Chemistry, Environmental Science (shown in the pattern) or Math, Engineering, and Computer Science (in black). About 30% of STEM majors identified themselves as FG. While women can be an under-represented group in some STEM disciplines, at our university, this does not seem to be true with the exception of the engineering program.

Figure 2. Descriptors of First-Year STEM majors by program of study

Attrition Is a Common Problem in STEM Majors Nationwide in 2013, about 28% of bachelor’s degree seeking students and 20% of associate’s degree seeking students initially pursue STEM degrees. However, only 52% of those bachelor’s degree seeking STEM students persisted to that degree, while 20% left the university without a degree (1). Moreover, as described in 2013, 31% of associate’s degree seeking students persisted to that degree. Thus, retention of STEM students is a significant concern. The call to produce more STEM graduates corresponds with the need for such professionals if we are to remain strong as a country; however such calls point to the barriers to raising that workforce, including the need to reform STEM higher education (2). While some argue that lack of interest and/ or background preparation is the reason for lack of retention, particularly of URM students, there seems to be little evidence that this deficit model is the reason that students leave STEM 111

(3). Instead, URM and first-generation students seem to be retained at lower levels than other students for a variety of complex and interacting factors (1–6). Reasons for this attrition are sometimes described as 1) cognitive (self-efficacy, grit), 2) contextual (climate, support, and barriers) and 3) cultural (self-identity and science identity) (6). While there may be differences between URM and FG students, studies on their retention have shown similar trends namely that these students are often retained in fewer numbers than their peers (4). Strayhorn (among others) has pointed out the significant and rapid increase in FG students attending college and some barriers that this student population faces (5). For example, first-generation students often have less social and cultural capital than their non-FG peers. Padgett notes that not all evidence-based practices work equally well for FG students (6). This study found that while FG benefit more than their non-FG peers from rigorous academic challenges, interactions with faculty, such as mentoring or research, might have negative impacts on FG students rather than the expected positive impacts. This study suggests that there is more to what is happening than simply adopting “high impact” practices (7) intended to increase interactions between faculty and students.

Figure 3. Retention Comparison STEM (first pair of bars in each set) vs. non-STEM (second pair of bars), URM (gray) vs. non-URM (black).

112

Most of the switching from STEM majors seems to occur within the first year, maybe two of the college experience (8). In this study, by the third year 59% of students had left the STEM program that they chose when in high school. The first year to second year transition is the major attrition point for all students at EMU as well. In Figure 3, we track and compare retention for first-year STEM majors (the first pair of bars in each section) and non-STEM majors (the second pair of bars). Historically (2011-15), URM STEM students (gray) leave both the STEM major and the university in higher numbers than the non-URM group (black). For the URM STEM bar in 2011-15, the hatched area on top indicates students that are retained at the university, but not within the major. Additionally while retention lags for URM non-STEM majors behind the non-URM non-STEM majors, there is still a difference between the URM STEM and the URM non-STEM majors. We also show in Figure 3, retention after the first year of our intervention (2016). The retention for URM students within STEM still lags significantly behind the non-URM population; however, we note a significant change in the URM students retained at the university (the hatched gray area). It’s too early to draw conclusions from these data.

Why Do Students Leave the STEM Major and What Interventions Work? In this section, we will explore some of the complicated reasons why URM and FG students might leave STEM majors. One significant factor that explains retention in STEM majors that are in the physical sciences (engineering, math, and computer science) is math preparation and student perception of their math preparation (9). Pre-college experiences (grades, success in advanced math and science classes) seem to be associated with higher success in college. Moreover, in this study, and others (10), first semester GPA was found to be a good predictor of student success. We do not discount the importance that pre-college preparation has on student success. However, success in math is not equally important for all STEM majors. For those STEM majors not in the physical sciences (such as biology), the perception of social fit is more aligned with retention. Support Services Such as Peer Tutors If academic success creates more success, then one way to improve STEM retention is strengthen academic success through student support services. There are many models for peer tutors, learning assistants, and peer mentors, but they share the common outcome that students who use these academic support services were more likely to graduate than those who used very few of these services (11). In one study, students in a large-enrollment, introductory biology course, who regularly participated in peer tutoring supports, had better course success than those who did not (12). For example, students which interacted with the peer tutors had improved exam grades and were more likely to persist in the course than students who were struggling but did not interact with peer tutors. The Learning Assistant program and the national learning assistant model established at the 113

University of Colorado, Boulder, has been shown to recruit future science teachers, improve content knowledge, and to shift how professors think about teaching and learning (13). Peer tutors often benefit academically from the tutoring experience (14, 15). The process of teaching the content often enhances their understanding of the disciplinary content. Being a peer tutor requires that the tutor review existing content knowledge, potentially reorganizing it or creating new connections, while trying to clearly and simply summarize content knowledge (13). Kuh names peer tutoring as one of the highest impact activities in which a student should engage (7). Participation in peer tutoring could reduce the gap between white and URM students, in that the gains in retention for Black and Hispanic peer tutors are greater than for white student (16). A study of peer led team learning (PLTL) in computer science courses found that peer leaders reported increases in self-efficacy and content knowledge (13). However, in this study, there was no difference between URM and majority students in these gains. These examples point to the importance of factors beyond STEM content for student retention in STEM. Grit or Persistence It can be difficult to determine exactly the mixture of reasons why one student persists and another stays. While there is a link between prior academic preparation and achievement, there are other personality variables, including grit, which could explain achievement (17). We often assume that persons who achieve do so because they demonstrate perseverance or grit. Grit specifically means working through challenges (continually, with sustained interest in the goal, despite failure or plateaus) and there are multiple ways of measuring this (18). However, when STEM programs study the link between these means of measuring grit and student retention, sometimes there is a connection between grit and success (19) and sometimes there is not (20). Lack of High Impact Practices One common criticism of higher education is a lack of pedagogical practices that inspire and retain students. Our programs use a number of ‘high-impact’ (7), evidenced-based approaches to engage students in science, such as 1) required research by all students 2) embedded research projects in traditional coursework (21) 3) active-based learning pedagogies in the classroom 4) writing intensive classes 5) community or service learning 6) internships and 7) a cross-cultural (global learning) requirement. Due to our retention numbers (Figure 3), we suspect that high impact practices alone are not sufficient to change retention in STEM. However, there is some evidence that high-impact practices do influence socio-cognitive factors such as science self-efficacy, grit, and/or science identity. Undergraduate research experiences are one such example of high-impact practices that have been linked to student success and retention. We have a number of research experiences embedded in our courses, such as a phytoremediation project in our general chemistry class (15). Course-based undergraduate research experiences or CUREs are easily adaptable to larger universities and have become 114

fairly common (16, 17). Studies on the longer-term impact of participation in a CURE, are complicated. For example, while one such iteration, the Freshman Research Initiative (FRI), did not significantly impact student GPA, students who completed all three semesters of the FRI were more likely to graduate with a STEM degree (22). There could be many reasons to explain these differences. A model developed by Corwin, et al. connects the research skill development with cognitive and cultural gains such as increases in self-efficacy, motivation for science, and persistence in the face of obstacles (23). The value of this model is that it names the many linkages among these different parameters that important for student retention in STEM.

Science Self-Efficacy and Identity The high impact practice of student research also has been shown to cause some students to identify strongly with science. Chemers suggests that science identity developed through research experiences is mediated by students’ science self-efficacy (24). Chemers uses Lave and Wegner’s situated learning perspective (25) to describe how students are developing an identity of themselves as scientists through joining a community of practice. In this apprenticeship model, we would anticipate that students develop a stronger identity as a scientist through the practice of science. However, if a student does not have a strongly developed sense of self-efficacy within the domain (in this case, science) in which they are doing research, they may not fully incorporate this domain (science) into their identity. A survey developed by Chemers to measure self-efficacy has been used in a number of studies on the success of students (26). A meta-analysis of gender differences finds similar gaps as is seen for FG and URM students (27). Namely, there appears to be little to no difference in academic performance early in their academic careers, but women, still report lower selfefficacy than men. Additionally, women express lower science identity, and these gaps do not change during their undergraduate studies. However, the cause of these gaps is not well understood. Students’ perceptions of self as a scientist might be tied to how useful such a STEM major seems to be in achieving a career goal (such as in the health sciences) (28). Identity development seems to be an important factor for motivation to study and persist in STEM. Science identity provides meaning for science experiences both by the individual participating in the science and for how society creates potential interpretations of individual actions (29). For women who were successful in STEM, this was evidenced in the importance of external recognition by others of the perception as the individual as a scientist. In other words, for women identity as a scientist was formed by both their passion for science and by external validation of their role as scientists. If a woman failed to receive this recognition, identity formation was disrupted which impacted academic achievement. These examples indicate some of the inherent complexity in separating student experiences from student cognitive and cultural variables to explain why a student is retained in STEM. 115

Socio-Cultural Reasons Our model for intervention assumes that attrition is a combination of sociocultural factors (30) related to the interactions between self (cognitive features such as self-efficacy or grit and cultural identity) and others (contextual and cultural cues about climate, support, barriers, and science identity) (6). While we do not discount other barriers to student success and retention, we suspect that many URM and first-generation students experience discomfort when encountering the null curriculum implicit in STEM. This implicit STEM culture or the null curriculum (31) is something that we do not teach, we rarely can even name for our students, and yet may be quite important to the discipline. For example, the null curriculum could be about thought processes (thought processes that are named as ‘intuitive’), content (math needed to succeed in chemistry) or affect (values or ethics). The implicit STEM culture may create barriers for students that are related to student self-identity. Brint, Caswell, and Henneman describe a cultural distinction between how STEM programs and non-STEM programs value engagement (32). In particular, in non-STEM (arts, humanities, social sciences) engagement is defined as interaction between students and professors, participation in and outside of class, and demonstration of interest in ideas. Whereas, for the STEM field, engagement is seen through quantitative skills (that improve with practice), collaborative problem solving, and a greater emphasis on financial award in future careers. They describe how students are ‘socialized’ into these values over the course of the career. However, if the student is asked to abandon their identity (that may differ from this implicit STEM identity) as they are socialized, they may not see how they ‘fit’ or may not want to fit in a STEM major. Aikenhead and others describe this experience as multiple border crossings (33). The student is entering the culture of our university (and those cultures are likely different school to school); they are also experiencing the ‘micro-culture’ of the first-year, large-enrollment STEM classroom, which may or may not be learner focused. In addition, they are encountering the culture of a reductionist and materialist science worldview. The culture of ‘science’ is not a monolith; the (sub)disciplines also have their own cultures. Thus, if a student is taking multiple STEM courses in their first semester on campus, they are crossing between many cultures, some or all of which may be new to them, and trying to navigate those differences while mastering content knowledge. Thus, the null curriculum could cause cognitive or cultural dissonance for students that hinders student identity formation. To address this issue, we also sought to provide additional supports to URM and FG students to help them navigate both the university and the discipline. First, we sought to use Lave and Wenger’s community of practice model (25) to create a learning community designed provide tools for STEM and writing faculty to be more culturally responsive in their teaching. Additionally, we began to modify our peer tutor program as peer tutors can function as the bridges between students and professors.

116

Research Questions How does participation in a professional learning community, shift faculty teaching identity, practices, and beliefs? How does a peer tutoring program focused on supporting all students – especially URMs and FG students – impact the tutors’ perceptions of identity and efficacy?

Methods Faculty as the Primary Focus of the Intervention Faculty interact with students in their academic spaces, set the classroom culture, make the assignments, decide the means of assessment, and thereby select the means by which academic success is measured, and also inadvertently set the means through which the URM students determine their own fit with the major, with the college, and/or with the higher education in general. Lasting change in the academy is driven most often by full-time faculty. While hiring URM faculty in the sciences must be a high priority, attracting URM faculty can be a challenge at mostly white colleges, in part because of some of the same cultural reasons why URM students might not be retained. Thus, we argue that ownership of the problem by all faculty in STEM is important if we aim to alter the STEM faculty culture (34).

Description of Faculty Intervention We designed a socio-cultural (24) intervention that asks STEM professors to adopt the identity of a teacher. Thus, we aimed our interventions at 1) naming and engaging the problem, 2) increasing the learning-centered pedagogies (35) in college STEM classrooms, and 3) helping faculty learn to effectively shift (36) their teaching identity, practices, and beliefs to a more learner-centered and culturally responsive stance. The current reality in most STEM classrooms is one that remains lecture-based and primarily uses teacher-centered instruction. When students fail to succeed the problem is viewed as lack of readiness or potential in the student. Even if active- learning strategies are adopted, they may or may not be learning-centered. This “fix-it” approach of looking for a quick fix strategy often defaults to perfectly logical explanations for student behavior, which often summarize, as “students are not ready for college.” Thus, we sought to name and engage the complexity of the problem before discussing best practices for evidence-based teaching for diverse student learner success. Implicit bias and deficient stereotyping is often subconscious and may be present even within race and gender groups (37). Thus we argue that instruction in pedagogical techniques without cultural context and understanding may not ultimately be successful. 117

Next, we took a longer approach to teacher development. The current reality is that professional development in pedagogy is often limited to attendance at conferences or single-event workshops, which may not result in identity shifts or significant change in practices. Faculty involved in a community of practice (25) are more likely to develop lasting changes in their beliefs, practices, and teacher identity. One of our underlying assumptions is that cultural dissonance will occur as part of this training because professors tend to teach from their own cultural identities. Professors find it challenging to form an identity as a teacher, in no small part due to the tension between a research identity long developed over the course of a STEM career and a teacher identity, which may or may not be highly valued at their institution or even broadly (38). The leadership team invited all STEM and writing faculty to participate in a Professional Learning Community (PLC) devoted to Diversity Responsive Teaching. While self-selection might be a potential problem here, we had sufficient resources to include our entire STEM faculty. The cohort had thirteen participants in year one, and an additional six are involved this year. The faculty members were STEM and writing instructors. Between the two cohorts, 88% of our STEM faculty will have participated in this intervention. The leadership team included first year writing faculty, also all full-time, as writing courses represent another barrier at our institution for retention. There remains one full-time STEM faculty member who did not elect to participate. A PLC provides structured exposure to new views of teaching or pedagogical strategies by someone with experience; thus the PLC was led by someone the teacher education program who designed the activities with input from our STEM leadership team. Additionally, facilitated learning experiences over time match the rate of the identity-developmental process for faculty. Finally, a PLC becomes a safe social network to do the ongoing reflective work of interpretation of what is happening in the classroom. The PLC was designed to address known barriers to professional teaching development such as time and motivation by scheduling hourly meetings once a month with short activities in between so that the time commitment was not intense but had sufficient duration to develop a community. Faculty who completed the work in the first year were given a small stipend and invited to take on a more intensive Scholarship of Teaching and Learning (SoTL) project that could lead to publication or presentation in the second and third year. After the yearlong PLC, 6 STEM faculty are being mentored in SoTL projects related to teaching and learning around URM retention. The goal of these workshops was to move faculty towards best practices for URM retention, based on reflective teacher and shared vision models. The trainings began with a daylong unconscious bias training for all faculty members at the university. This was followed by another intensive (half-day) training in August, before classes started. In this intensive training, faculty members were provided resources on cultural aspects of learning and identity, growth mindset and deficit models. The workshop provided an introduction to the research surrounding the needs of URM and FG students, combined with reflective practice in which professors examined themselves, both their own culture and how it has shaped their teaching. The workshop also introduced the ways in which culture is 118

largely implicit, and how much of our behavior as humans is based on parts of our culture which remains hidden even to ourselves. Faculty were asked to consider ways in which they might utilize the first weeks of the semester to implement initial, simple culturally responsive practices that would invite URM and FG into a deeper engagement in the course and discipline. During the school year, we met face-face as a “class” once a month. All of the inputs involved active components, which modeled effective pedagogical practices. In these meetings, members completed an activity, provide rapid written reflections, and reflected in small groups. Sample course topics include: • • •

Perceptions of faculty behavior by students of color Imposter syndrome (first-generation and other students) Active learning strategies with improved scaffolding or learning supports and how this leads to greater student success (in particular for URM and first-generations students).

In addition to inputs in the form of videos, presentations, lectures, and role-playing, there were several activities we completed individually and processed as a group. Sample course assignments for the faculty were written reflections and peer review through an online course management system. We also completed the Intercultural Development Inventory (IDI) (39), which helps faculty understand their cultural beliefs. Other examples of assignments include: • •

Application of Understanding by Design (40) to redesign a lesson with active student learning/participatory lectures Peer Syllabus review and re-design for clarity, capacity-language (vs. deficit language), culturally responsive language

Through these activities, faculty develops accountability to one another, and the learning curve for faculty as teacher is normalized. The variety of contemplative and pragmatic activities help faculty to recognize enculturation of themselves as teachers and perhaps develop an understanding of what is happening for their students in their science identity development. The more diffuse nature of the PLC activities gives faculty space to incrementally test out intentional shifts in practices. Peer-Tutoring at EMU Historically, we have had several types of peer-tutors on campus, and diverse offices managed them. The support and training of the peer tutors has been inconsistent. Generally, there have been two types of STEM peer-tutors: 1) those who worked for the Academic Success Center (located in our library) and 2) those who enrolled in an independent study related to tutoring/ teaching in a specific class, often with the goal of earning community or service learning credit (supervised by an individual faculty member). Neither of these types of tutors was trained in an official capacity. Only the peer-tutors in the writing program have a required training before they could work as a tutor. All tutors who worked 119

through the Academic Success Center participated in an initial orientation and were invited to occasional luncheons for sharing and troubleshooting. Typically tutors had limited to modest opportunities for contact with each other or the faculty of courses that they supported. Finally, our peer tutor demographics were more reflective of our faculty demographics than of our students. Unlike rapid shifts in tenured faculty, it is easier to pay attention to the demographics of the peer tutors and to work to have them reflect the demographics our students. We redesigned the peer tutor program to address the challenges described above. While peer tutoring can be effective, our decentralized approach did not match best practices for optimal success. We suspected that peer tutors would benefit from formal, “start-up” training on how to support students of diverse backgrounds and learning styles. Moreover, like our faculty development project, we planned to provide follow-up during and after the semester to provide space for peer tutors to continue to reflect and learn. A pilot program (41, 42) at EMU looked at the effects of peer tutors “embedded” in courses. Some of the key recommendations of these programs were to connect the peer tutors closely to faculty for feedback. Thus, we sought to “embed” whenever possible all STEM tutors with a faculty member or a course. Peer tutoring will likely be most successful in facilitating improved URM student outcomes if faculty is heavily engaged with the peer-tutor program.

Description of Peer Tutor Intervention We sought first to change our recruitment process. We paid careful attention to the current demographics of successful peer tutors that we wanted to retain and compared those to the demographics in our classes. We created lists of URM and first-generation students who had been successful in the STEM classroom and contacted all of them to probe their interest in becoming a peer tutor. Many students self-reported low STEM self-efficacy in response to the tutoring invitation despite their prior academic successes. In the first year, we hired seven (nine were invited) additional tutors as part of the NSF grant increasing the total number of STEM tutors to 22; five of the additional seven were URM and/or first-generation students. Tutor training was offered to all tutors. The formal training of peer tutors included several aspects that paralleled faculty training: •

General tutoring guidelines (from prior trainings) ◦ ◦

• •

Recommendations for how to tutor Learning styles inventory

Teaching to diverse learners training (new, related to Faculty training) STEM discipline- (or professor-) specific (new) ◦ ◦

General expectations for STEM disciplines Interviews with a STEM professor 120



Specific information about how peer tutors can support learning in the classroom

This included roughly 8 hours of work, spread out over in-person and online formats. To begin, students were asked to complete a learning style inventory instrument (43) online and followed by a guided reflection on their results in an online forum. Additionally, they participated in online forums for introductions, discussions about past learning experiences, and a common reading of the most current research findings on effective tutoring. Next, there were two 1-hr interactive lunch sessions to cover 1) trends, needs, and issues surrounding URM students and 2) a similar format on first-generation students. As part of these luncheons the students were exposed to various scenarios so that they could practice their skills. These initial trainings were enhanced during the semester by a series of nine online discussion forums facilitated by their supervisors. The overarching goals of the training were to 1) increase awareness of the peer tutors about the needs of diverse learners and 2) discuss how tutors can leverage this awareness in practical ways. A related goal was to build a community among the tutors. We did this through interactive introductions and some games, followed by tutor lunches. We wanted to continue to test the results of embedded peer tutors in the STEM classroom as seen in the pilot program (39, 40). Thus, we sought to connect peer tutors with faculty. We approached both peer tutors and faculty to check their interest and availability. We were able to match 10 of the 22 STEM peer tutors with specific courses. These ten tutors worked closely with seven different STEM faculty. As part of their peer tutor training, STEM tutors interviewed a STEM faculty member to learn more about how the professor structures their course. Initial understanding of peer tutor gains is being assessed with surveys and focus groups. As we only have results from year 1 of a three-year process, this chapter will provide preliminary findings that are shaping the project.

Preliminary Results of Faculty Training This Is Hard Work As others have found before us (44, 45), the process of having a professor study their teaching and how that impacts learning in their classroom is difficult, in part because it asks the professor to become vulnerable. Using Aikenhead’s description, a border crossing occurs as faculty begin to study their teaching and how that impacts learning in their classrooms. In addition to time, it requires “an immersion in a different intellectual language and culture, experiential learning, personal reflection and an iterative process of moving backwards and forwards between the familiar STEM approach and a different way of thinking” (42). Thus for a group of faculty to engage in this sort of professional development, the faculty must also comprehend and internalize the null curriculum in the Scholarship of Teaching and Learning (SoTL). For example faculty described the difficulty of managing both rigor through high expectations with intentional scaffolding for support in this way: “You can 121

have both high expectations and high support. But it’s hard to do day after day, in a cohesive manner, with a text written by someone else.” Simmons et al describes how a group of scholars have managed the conflict inherent in new identity development and how they worked to resolve the conflict between their primary discipline and their SoTL discipline identity (43). They describe this process as liminal space where deep reflection can occur. The value of liminality as a metaphor is that it is fluid and transforming, without immediate or clear closure. One frequent request from our faculty for a resolution: which pedagogical tool “worked” for all students? For example, in the September forums (note: this is early in the intervention), one faculty member wrote: “I know we are supposed to be ok with non-closure, but I really would like some suggestions from more experienced instructors on how to deal with such a pattern!” There was frequent frustration when a professor would try something new in the classroom, and there would not be immediate success, or it would not work for some portion of the classroom. For example, in December (mid-year of the intervention) one faculty member wrote this about implementing a change in their class: “10% still had several completely wrong answers, even after think-pair-share, whole group discussion, and me actually writing down [the correct] answers. I’m not sure where this takes me.” It was tempting to blame either the student or the tool. As seen in prior studies, faculty, even STEM trained scientists, tend to believe their experiences over empirical based studies (46). Thus, faculty sometimes struggled to integrate their experiences with the instruction they were receiving in the PLC. After being asked to review and revise the syllabus to be more invitational for URM and first-generation students, one professor reflected: “[I have] always designed my syllabi around what I appreciated as a student: I wasn’t as worried about the granular details. And to be honest, I’m still not sure how much of the "why" should be in a syllabus.” Thus, we saw similar features in our professional learning community as Simmons et al describes: 1) doubt and insecurity (which we will describe as loss of professor efficacy) 2) risk taking (willingness to try new pedagogies and reflect on them) 3) working towards assimilation (trying to manage empirical studies, their experiences, wanting a resolution to conflict) (44). This cycle of reflective inquiry is a natural product of identity development as a teacher.

Deficit vs. Growth Mindset Additionally, we noticed some shifts in beliefs from the beginning to the end of the intervention about student deficit vs. growth mindset. In August, the reflections included numerous descriptions of observable behaviors of URM students such as not asking for help, being late or missing class, and not taking advantage of faculty office hours and tutoring. When phrased as “Students are not doing x”, the implied onus is on the students, and this tends to reflect a “deficit” rather than a capacity or growth mindset. Faculty would say things such as “perhaps there is a subset of the class that just does not understand the expectations for college.” 122

This was also framed as an either or between rigor or high expectations and support or learning scaffolds. A number of professors when asked at the beginning of the professional development course what their hard question (intractable problem) was named this: “how (or if) to develop different materials that aren’t watered down, and retain all the academic rigor that I would normally use.” This question reflects a deficit model and a lack of ‘faith’ that changes in the classroom could maintain academic rigor and improve student success. Another professor put it this way: “My "hard question" is an old and tired one: how do we modify assignments and expectations to support specific students and groups of students without "dumbing down" those assignments and expectations?” We did see positive shifts in this thinking over the course of the intervention. In the later reflections, faculty were more likely to say things such as: “High expectations [are] important, standards are important…where there is a disconnect, there must be assessment to discern missing links – followed by support…and not always by instructor.” We see in this phrase that “disconnects” (i.e. learning is not happening) triggers a reflection on what is happening, without an immediate jump to a conclusion. This statement also reflects a shared responsibility for learning in the phrase “not always by the instructor.” It could be that the instructor needs to make a change or it could be something for the student or perhaps other supports (peer tutors) could be more successful. Towards the end of the year, faculty were more likely to say things that reflected a growth mindset, such as: “I think what this gets at is person-to-person, human interactions, promotes a better atmosphere for learning.” Doubt, Insecurity and Efficacy as a Teacher This shift about ownership of the problem as described above seemed to parallel a shift in teacher efficacy. Earlier in the intervention, faculty were more likely to say things like “this is just the way it is, there’s not much that I can do about it” or to want to find a magic tool that worked in all situations, for all professors, for all students, the first time it was used. “[I] have tried to restructure the course to cover learning and study strategies, but that does not seem to have an effect on performance in the first exam.” As the year progressed, there was what we describe as an increase in professor efficacy, where the faculty member would try something or would like to try something but rationale was lacking for why it worked (or did not). For example, “I want to use ... case studies as a better mechanism for drawing students into a discussion of the content. Think-pair-share is a useful means of doing more of this, but I struggle with finding questions are worth discussing…I tried, and feel like I failed, to put that into something that inspires much discussion.” This represents a willingness to take a risk (see below), but also some lack of efficacy as a teacher. The professor is not sure that the work of changing the class is worth it and if the pedagogical tool can be effective at the goal (engaging students). In May, we found more professors able to describe the rationale for using certain pedagogy, and they were more likely to believe in the effectiveness of making changes. “I found that I need to be more intentional in designing problems that leads towards higher order thinking (both towards the “next” thing in the 123

text and Bloom’s higher order) AND in managing my own responses to those activities.” Statements such as this one also reflect a belief that scaffolding for support is important in student learning. Risk-Taking: Willingness To Use Learning Centered Strategies Even faculty who displayed resistance to change in their writing participated in the activities and the reflection. For example, looking at three data points over the course of the intervention (September, December, and May), 7 of the 12 faculty members participated in all of the prompts. For those individual months, 10 of 12 responded in September and May, and 8 of 12 in December. This speaks to faculty willingness to participate and to take risks. Although faculty demonstrated willingness to engage by attempting activities and reflecting upon them, they were initially resistant or skeptical towards the effectiveness of the work. When asked to reflect on the syllabus review, one faculty member said “This activity was much more helpful than I thought it might be. My initial skepticism stemmed from the fact that I think the syllabus is essentially a technical document that students rarely (if ever) read.” Throughout the intervention, there were some faculty who retained this “surprise” at the usefulness of the task, but some faculty also demonstrated a stronger willingness to engage. For the same assignment, another faculty reflected in this manner: “my other assignments are much more ambiguous in terms of how I evaluate them. I rewrote my grading practices throughout my syllabus to clarify exactly how I plan to evaluate each.” What we would like to understand is how professor identity and efficacy impact the professor willingness to engage in these activities. Working towards Resolution Finally, faculty who participated in this intervention used the reflective writing process as a means to find resolutions between the tensions they were experiencing. For example in one writing sample, an individual faculty member expressed these ideas: ◦ ◦ ◦

I have a vision for how the learning process “should” go in each class, but I always have students who don’t believe my vision. How much effort should I invest in reaching out to that second group? Every student has a story, and there’s no algorithm that would allow me to connect with, and listen to, each students’ story.

In these three phrases we can see the faculty member attempting to discern who “owns” learning, some frustration with trying to reach each student, and some sense that adaptable pedagogies are needed. The process of writing seems to have been an important step in the resolution of tensions. As one faculty member stated at the end of the intervention: “The learning process cannot be set up [as] an immovable plan and [be] expected to move forward without adjustment.” 124

Preliminary Results of Peer Tutor Intervention Peer tutors also demonstrated a willingness to participate in both the trainings and the reflective activities afterwards. The following results reflect participation of all peer tutors, both those who are participating voluntarily and those funded by the NSF. Thus the response rate might be skewed by student commitment to the project. In the fall semester there were nine prompts for reflective postings, with an average response rate of 61%. Almost all students participated at least once (87% each semester). We see participation in reflective postings significantly drop-off around midterms and finals as this is a demanding time for students. In response to student feedback about the frequency of prompts, there were fewer prompts in the spring (six), but the average response rate dropped to 54%. We also surveyed peer tutors about which aspects of tutoring and training they found most valuable. As seen in Figure 4, peer tutors self-reported that the learning-style inventories, the interactions with the STEM professors, and the training on student-centered learning was more important to their work as peer tutors. Follow-up discussions with the tutors suggested that their low ratings of the importance of working with first-generation students and cultural awareness did not imply that they saw these areas as unimportant; they simply felt that these areas were more abstract. Several tutors also mentioned how they lacked confidence on how to implement specific approaches with URMs or first-generation students in an effective way.

Figure 4. What trainings peer tutors described as most valuable One important goal was to “embed” peer tutors in courses. Seven fall courses and eight spring courses had such a connection clearly identified. Tutors, faculty, and supported students all valued this connection. This was supported by their comments in surveys. For example, a peer tutor described the value of working closely with a professor in this way: “It was useful knowing what students have struggled with in the past so that I can focus more on those areas for study sessions. It was also helpful to learn that the professor is willing to work with students when they come to her… I can be the mediator between the student and professor in helping the student get to the professor without it feeling so intimidating.” This role 125

of the bridge or mediator between professor and student was one that we hoped both the tutors and students would recognize. For the learning style inventory, peer tutors were able to make connections between what they would do as a tutor and the training itself. For example, a tutor said “It was very informative to learn about the different types of learners. Before this, I knew there were different ways of learning, but I didn’t know what they were or how to more effectively teach these people.” Additionally, we had hoped that the peer tutors would see each other as a “Learning community”, and this appears to be the case. Another peer tutor said “Since this is my first time ever tutoring I think it’s helpful to be able to have a community of tutors to know and be able to turn to when I need help. I also have made a few friends along the way!” As we had specifically increased representation of URM/ first gen students in our peer tutors, we were curious how this experience would impact their view of themselves. One tutor described the experience in this way: “Being an individual that fits into both categories [URM and first generation], I found that the discussions focused on these topics helped me a lot in understanding where I stand and also how I can successfully help individuals in the same situations.” Unfortunately, while student perception of self improved with regard to their efficacy as a tutor, these students continued to self-report low identity in STEM. Based on a weighted ranking scale, the peer tutors tended to see themselves as a “good helper” (5.5), a “good student” (5.3), and “an empathetic person” (4.9) moreso than as a “good STEM student” (3.3), a “good communicator” (3.3), or a “good scientist/mathematician” (2.4). Finally, we note that some of what the peer tutors describe mirrors what we see for faculty involved in these interventions. Peer tutors and faculty describe an increased awareness of self and others. A peer tutor described it like this: “I try and think about their point of view and where they are coming from. I have to remind myself that I do not know their whole story and background that can significantly impact them.” Additionally, both peer tutors and faculty report gains in self-efficacy. For one peer tutor, they described it this way. “I feel much more confident in communicating with people I don’t know and I can hold conversations with students about classes and interact with them in a way that is beneficial to their learning.” Another tutor felt she had “grown because tutoring has given me more perspective in what students need when they are struggling.”

Limitations and Conclusions It is too early in our project to say anything concrete about STEM retention, although we do report the data from the first year (Figure 3). Moreover, finding changes in peer tutors and faculty identity, practices and beliefs does not prove causation for changes in retention and attrition. We are continuing to study retention and enrollment data for URM and FG college students in STEM and looking at other success data (such as first semester and first year GPA). This data will be matched with surveys of interest and understanding of intended majors. Finally, we will follow-up with interviews regarding intended majors and issues 126

they have faced in their first year. We hope to connect this information with the studies we have described here. Our early qualitative findings match the literature in that professional learning communities seem to be valuable for developing faculty and peer tutor identity and shifting beliefs. Our faculty demonstrated both an understanding of their ownership of the problem of attrition and some increased efficacy on how to increase student success. This intervention seems to have impacted professor identity in the domain of teacher, in that six faculty agreed to undertake scholarship projects in the area of teaching and learning. Additionally, intentional training of peer tutors was successful in connecting faculty with the tutors and seems to have increased the tutors sense of efficacy as a tutor.

Acknowledgments This work was made possible by funding from the National Science Foundation, Division of Undergraduate Education, award number: 1611713. This project was generated by conversations, ideas, and planning from colleagues who are now at other organizations: Dee Weikle and Susannah Lepley. We are grateful for EMU student peer tutors, EMU STEM students, and faculty workshop participants. We also acknowledge the many contributions of the director of the Academic Success Center, Linda Gnagey. We are grateful to our external evaluator, Carol Hurney (Center for Teaching and Learning, Colby College), who has shaped our understandings.

References 1.

2.

3.

4.

5. 6.

Chen, X. STEM Attrition: College Students’ Paths into and out of STEM Fields. Statistical Analysis Report; NCES 2014-001; National Center for Education Statistics: 2013. Olson, S.; Riordan, D. G. Engage to Excel: Producing One Million Additional College Graduates with Degrees in Science, Technology, Engineering, and Mathematics; Report to the President; Executive Office of the President: 2012. Byars-Winston, A.; Estrada, Y.; Howard, C. Increasing STEM Retention for Underrepresented Students: Factors that matter; Research Brief; Center on Education and Work, University of Wisconsin at Madison: Madison, WI, 2008. Ishitani, T. T. Studying Attrition and Degree Completion Behavior Among First-Generation College Students in the United States. J. Higher Educ. 2006, 77, 861–885. Strayhorn, T. L. Factors Influencing the Academic Achievement of FirstGeneration College Students. NAPSA J, 2006, 43, 82–111. Padgett, R. D.; Johnson, M. P.; Pascarella, E. T. First-Generation Undergraduate Students and the Impacts of the First Year of College: Additional Evidence. J. College Stud. Develop. 2012, 53, 243–266. 127

7.

8. 9.

10. 11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

Kuh, G. D. High-Impact Educational Practices: What are They, Who has Access to Them, and Why They Matter; Association of American Colleges & Universities: Washington, DC, 2008. Shaw, E. J.; Barbuti, S. Patterns of Persistence in Intended College Major with a Focus on STEM Majors. NACADA J. 2010, 30, 19–34. Dika, S. L.; D'Amico, M. M. Early Experiences and Integration in the Persistence of First‐Generation College Students in STEM and non‐STEM Majors. J. Res. Sci. Teach. 2016, 53, 368–383. Whalen, D. F.; Shelley, M. C., II. Academic Success for STEM and nonSTEM Majors. J. STEM Educ.: Innovations Res. 2010, 11, 45. Grillo, M. C.; Leist, C. W. Academic Support as a Predictor of Retention to Graduation: New Insights on the Role of Tutoring, Learning Assistance, and Supplemental Instruction. J. College Stud. Retention: Res., Theory Practice 2013, 15, 387–408. Batz, Z.; Olsen, B. J.; Dumont, J.; Dastoor, F.; Smith, M. K. Helping Struggling Students in Introductory Biology: a Peer-Tutoring Approach that Improves Performance, Perception, and Retention. Cell Biol. Educ. Life Sci. Educ. 2015, 14, 16. Otero, V.; Pollock, S.; Finkelstein, N. A physics department’s role in preparing physics teachers: The Colorado learning assistant model. Am. J. Phys. 2010, 78, 1218–1224. Topping, K. J. The Effectiveness of Peer Tutoring in Further and Higher Education: A Typology and Review of the Literature. Higher Educ. 1996, 32, 321–345. Hug, S.; Thiry, H.; Tedford, P. Learning to Love Computer Science: Peer Leaders Gain Teaching Skill, Communicative Ability and Content Knowledge in the CS Classroom. Proceedings of the 42nd ACM Technical Symposium on Computer Science Education, Dallas, Texas, March 9-12; ACM: 2011, pp 201−206. Finley, A.; McNair, T. Assessing Underserved Students’ Engagement in High-Impact Practices; Association of American Colleges and Universities: Washington, DC, 2013. Duckworth, A. L.; Peterson, C.; Matthews, M. D.; Kelly, D. R. Grit: Perseverance and Passion for Long-term Goals. J. Pers. Soc. Psych. 2007, 92, 1087–1101. Duckworth, A. L.; Quinn, P. D. Development and Validation of the Short Grit Scale (GRIT–S). J. Pers. Assessment 2009, 91, 166–174. Jaeger, B.; Freeman, S.; Whalen, R.; Payne, R. Successful Students: Smart or Tough Presented at 2010 Annual Conference & Exposition, Louisville, Kentucky, June 20, 2010. Bazelais, P.; Lemay, D. J.; Doleck, T. How Does Grit Impact College Students’ Academic Achievement in Science? Eur. J. Sci. Math. Educ. 2016, 4, 33–43. Cessna, S. G.; Kishbaugh, T. L.; Neufeld, D. G.; Cessna, G. A. A Multiweek, Problem-Based Laboratory Project Using Phytoremediation To Remove Copper from Soil. General Chemistry Labs for Teaching Thermodynamics and Equilibrium. J. Chem. Educ. 2009, 86, 726–729. 128

22. Rodenbusch, S. E.; Hernandez, P. R.; Simmons, S. L.; Dolan, E. L. Early Engagement in Course-based Research Increases Graduation Rates and Completion of Science, Engineering, and Mathematics Degrees. Cell Biol. Educ. Life Sci. Educ. 2016, 15, 1–10. 23. Corwin, L. A.; Graham, M. J.; Dolan, E. L. Modeling Course-based Undergraduate Research Experiences: An Agenda for Future Research and Evaluation. Cell Biol. Educ. Life Sci. Educ. 2015, 14, 1–13. 24. Robnett, R. D.; Chemers, M. M.; Zurbriggen, E. L. Longitudinal Associations Among Undergraduates’ Research Experience, Self‐efficacy, and Identity. J. Res. Sci. Teach. 2015, 52, 847–867. 25. Lave, J.; Wenger, E. Situated learning: Legitimate Peripheral Participation; Cambridge University Press: Cambridge, UK, 1991. 26. Chemers, M. M.; Hu, L. T.; Garcia, B. F. Academic Self-efficacy and Firstyear College Student Performance and Adjustment. J. Educ. Psychol. 2001, 93, 55. 27. Eddy, S. L.; Brownell, S. E. Beneath the Numbers: A Review of Gender Disparities in Undergraduate Education Across Science, Technology, Engineering, and Math Disciplines. Phys. Rev. Phys. Educ. Res. 2016, 12, 020106. 28. Perez, T.; Cromley, J. G.; Kaplan, A. The Role of Identity Development, Values, and Costs in College STEM Retention. J. Educ. Psych. 2014, 106, 315–329. 29. Carlone, H. B.; Johnson, A. Understanding the Science Experiences of Successful Women of Color: Science Identity as an Analytic Lens. J. Res. Sci. Teach. 2007, 44, 1187–1218. 30. Lemke, J. L. Articulating Communities: Sociocultural Perspectives on Science Education. J. Res. Sci. Teach. 2001, 38, 296–316. 31. Flinders, D. J.; Noddings, N.; Thornton, S. J. The Null Curriculum: Its Theoretical Basis and Practical Implications. Curriculum Inquiry 1986, 16, 33–42. 32. Brint, S.; Cantwell, A. M.; Hanneman, R. A. The Two Cultures of Undergraduate Academic Engagement. Res. Higher Educ. 2008, 49, 383–402. 33. Aikenhead, G. S.; Jegede, O. J. Cross-cultural Science Education: A Cognitive Explanation of a Cultural Phenomenon. J. Res. Sci. Teach. 1999, 36, 269–287. 34. Chasteen, S. V.; Wilcox, B.; Caballero, M. D.; Perkins, K. K.; Pollock, S. J.; Wieman, C. E. Educational Transformation in Upper-division Physics: The Science Education Initiative Model, Outcomes, and Lessons Learned. Phys. Rev. Phys. Educ. Res. 2015, 11, 020110. 35. Marbach-Ad, G.; Rietschel, C. H.; Saluja, N.; Carleton, K. L.; Haag, E. S. The Use of Group Activities in Introductory Biology Supports Learning Gains and Uniquely Benefits High-achieving Students. J. Microbiol. Biol. Educ. 2016, 17, 360. 36. Borrego, M.; Henderson, C. Increasing the Use of Evidence‐based Teaching in STEM Higher Education: A Comparison of Eight Change Strategies. J. Eng. Educ. 2014, 103, 220–252. 129

37. Burrell, M. J. O.; Fredericks, M. A. C.; Fleming, L. N. Succeeding as Engineering Majors: Cultural Ecology Theory and Perceptions of Within-Race Gender and Ethnicity Differences in Engineering Skills and Work Ethic. Presented at 2013 ASEE Annual Conference & Exposition, Atlanta, Georgia, June 23−26, 2013. 38. Brownell, S. E.; Tanner, K. D. Barriers to faculty pedagogical change: Lack of training, time, incentives, and tensions with professional identity? Cell Biol. Educ. Life Sci. Educ. 2012, 11, 339–346. 39. Hammer, M. R.; Bennett, M. J. The Intercultural Development Inventory: An Approach for Assessing and Building Intercultural Competence. Contemporary Leadership and Intercultural Competence; Moodian, M., Ed.; Sage: Thousand Oaks, CA, 2009; pp 203−217. 40. Wiggins, G. P.; McTighe, J. Understanding by design; ASCD: Alexandria, VA, 2005. 41. Weikle, D. A. More Insights on a Peer Tutoring Model for Small Schools with Limited Funding and Resources. J. Comput. Sci. Colleges 2016, 31, 101–109. 42. Weikle, D. A. A Peer Tutoring Model for Small Schools with Limited Funding and Resources. Proceedings of the 46th ACM Technical Symposium on Computer Science Education, Kansas City, Missouri, March 04-07, 2015; ACM: 2015; pp 514−513. 43. McCarthy, B. Using the 4MAT System to Bring Learning Styles to Schools. Educ. Leadership 1990, 48, 31–37. 44. Kelly, N.; Nesbit, S.; Oliver, C. A Difficult Journey: Transitioning from STEM (science, technology, engineering and mathematics) to SoTL; Center for Teaching, Learning, & Scholarship at Georgia Southern University: Statesboro, GA, U.S.A., 2012. 45. Simmons, N.; Abrahamson, E.; Deshler, J. M.; Kensington-Miller, B.; Manarin, K.; Morón-García, S.; Oliver, C.; Renc-Roe, J. Conflicts and Configurations in a Liminal Space: SoTL Scholars’ Identity Development. Teach. Learn. Inquiry: ISSOTL J. 2013, 1, 9–21. 46. Andrews, T. C.; Lemons, P. P. It’s Personal: Biology Instructors Prioritize Personal Evidence over Empirical Evidence in Teaching Decisions. Cell Biol. Educ. Life Sci. Educ. 2015, 14, 7.

130

Chapter 8

Studio Format General Chemistry: A Method for Increasing Chemistry Success for Students of Underrepresented Backgrounds Jane Brock Greco* Department of Chemistry, Johns Hopkins University, 3400 N. Charles St., Baltimore, Maryland 21218, United States *E-mail: [email protected]

In the studio model, the lecture and the laboratory are fully integrated. In our implementation, the class met for three 1 hour 50 minute periods weekly. Classes could include laboratory activities, interactive lectures or group problem solving. The course was targeted to students who were academically under-prepared to succeed in general chemistry. The course consisted of a mixture of students who were encouraged to enroll due to low chemistry knowledge on a pretest and open enrollment students. Regression analysis on students’ demographic characteristics suggests that students enrolling in the studio class had a 0.58 increase in chemistry GPA compared to those in the traditional lecture/lab combination course. Students co-enrolled in a traditional biology class also demonstrated greater improvement on the chemistry related questions on the biology final exam, and these gains appeared to continue in the second semester when all students took the same lecture/laboratory combination.

Currently over 60% of students who enter college with the intention to graduate with a STEM degree fail to complete it. This situation is even more problematic for women and underrepresented minorities, who make up over 70% of college students but only 45% of STEM degree recipients (1). In order to increase the number of STEM graduates, it is necessary to both improve

© 2018 American Chemical Society

retention rates in STEM through higher quality teaching, and specifically target interventions to under-represented students. A meta-analysis of studies published on active learning found that on average active learning techniques increase exam scores by 6%, and that students were one and a half times more likely to fail in more traditional courses (2). Active learning is beneficial to all students, but a highly structured active learning class in biology (3) and peer led team learning instead of a lecture in chemistry (4) have been shown to reduce achievement gaps. As introductory chemistry is required for many different STEM majors, success in general chemistry is frequently required to continue as a STEM major. Many studies have been conducted to examine the factors that influence success in general chemistry (5). Math SAT scores (6, 7), and prior chemistry content knowledge (8), have been linked to success in general chemistry. The very high correlation between pretest score and chemistry grade indicates that prior knowledge is important for success in chemistry (8). A comprehensive review of factors influencing general chemistry found that when high school preparation was included, that the effect of underrepresented minority status was quite small (5).They suggested that the studies that find a larger performance gap for underrepresented minority students do not control for the fact that underrepresented minorities frequently attend high schools where academic opportunities are more limited. The studio format of teaching involves integrating the lecture and the lab component of a science course in an active learning environment. The largest implementations of the studio model have been through SCALE-UP (Student Centered Active Learning Environment with Upside Down Pedagogies) (9), and many of the adopters have been instructors of introductory physics courses. The implementation for introductory chemistry has been more limited. Published accounts of studio chemistry include implementations at Rensselaer Polytechnic Institute (10), University of Michigan (11) and California Polytechnic (12). Many of these reports have focused primarily on how the model was implemented rather than extensive analysis of the success of this model. More recently, the group at Cal Poly has published a detailed analysis of the success of their studio chemistry model, and included data on the performance of their least prepared students (13).

Study Motivation and Research Questions Prior research has suggested that active learning techniques can serve to reduce the achievement gap. Pre-requisite knowledge is important to success in chemistry and many students from under-represented backgrounds have not been exposed to the prerequisite skills and knowledge during the course of their high school careers. The research questions are (1) What is the effect of a studio chemistry format course on student success, as measured by overall course grades, for students from at risk backgrounds (first-generation college students, low income students, and under-represented minorities)? 132

(2) What is the effect of a studio chemistry format course on a conceptual understanding of chemical concepts? (3) Can the understanding of chemical concepts be transferred to other courses?

Methods Description of Studio Chemistry Course Format The studio chemistry course was designed as an alternative to the traditional general chemistry lecture and lab course. The traditional model consists of three fifty-minute lectures and an independent weekly three-hour laboratory course. The lecture course is taught in two sections with two different instructors and has approximately three hundred students in each section. The laboratory course is taught in six sections of one hundred students by a single instructor. There is an optional peer led team learning program, and students with a weak background in chemistry are encouraged to enroll in chemistry with problem-solving, which is a 0 credit, 2-hour weekly course of approximately 20 students that provides students that opportunity to meet in a small group with a third chemistry faculty member to work on problem solving skills. The lecture sequence covers stoichiometry, chemical formulas, Lewis structures, VSEPR, Gases, Intermolecular Forces and Liquids, Enthalpy and Entropy, Vapor Pressure and Phase Diagrams, Gas Phase Equilibrium reactions, Acid-Base Chemistry, and Solubility. The laboratory portion of the course focuses on these topics and covers significant figures, calibration curves, and laboratory skills, including the use of volumetric glassware and filtration. The studio course was designed to provide an alternative to the lecture and lab sequence during the first semester only. As students would be taking the second semester traditional lecture and lab there was no flexibility in the level of coverage of topics. The studio course met for three 110-minute periods weekly. These meetings could take place in a classroom with movable desks or in a laboratory. Based on the available laboratory space for this project, the course size was limited to 32 students, and 29 students persisted in the course after the drop deadline for courses. This is consistent with the standard variation of students choosing classes and lab sections during the first several weeks of classes. The class time was used for interactive lectures (utilizing the i-clicker classroom response system) (14), group problem solving, demonstrations and laboratory activities. The laboratory activities were placed where appropriate to directly correspond to the material that was covered in lecture. All courses used the same textbook (Atkins, Chemical Principles) and the Sapling system for electronic homework. In the studio course, students were allowed two unexcused absences and were required to complete a make-up assignment to catch up from an excused absence. Instead of weekly homework, assignments were due in the Sapling system three times a week to make sure that students completed the necessary problems after each lecture to prepare for the next lecture. 133

Student Recruitment The course was listed on the enrollment software and presented to all incoming freshman as an alternative to taking both the corresponding lecture and laboratory course. The course was presented to freshman advisors as being focused for students who were underprepared as an alternative for students who would be taking the problem-solving course. Students who are considered at risk include students who are first-generation college students, Pell grant recipients or otherwise low income, students from under-resourced high schools and under-represented minorities. These students are invited to participate in enrichment programs through the Center for Student Success. Students in these programs are tested for chemistry knowledge during the summer before matriculation through the ALEKS (15) program and strongly encouraged to complete summer online coursework in this system. Students who were identified as having weak chemistry knowledge, approximately the lowest third of incoming chemistry knowledge based on testing the entire freshman class in Fall 2014, were strongly recommended to take either the problem-solving course or the studio chemistry course. The 30 students recommended for intervention divided themselves equally between the studio course and the traditional lecture/lab sequence. One third of the students in the lecture/lab sequence chose to take the additional problem-solving course.

Data Collection Math SAT score or ACT scores, AP chemistry scores when available and course grades were obtained from the registrar’s office. ACT scores were converted to SAT scores using a correspondence table, and the maximum of the math testing was used. ALEKS scores and Exam 3 scores were obtained from the faculty members records. A pre-course survey consisting of parts of the chemical concept inventory and parts of the Colorado Learning Attitudes about Science Survey-chemistry version were given during the first class for course planning purposes. The data analysis was approved by the institutional IRB. At the end of the class, the chemical concept inventory and the Colorado Learning Attitudes about Science Survey were repeated, and students signed a consent before taking these assessments. Surveys of students under the age of 18 were eliminated.

Results and Discussion Student Demographics A total of 66 students were administered a pretest for chemistry and completed either the standard general chemistry sequence or the studio chemistry course. 134

The students who were administered the pretest were all first-generation college students, Pell grant recipients, or under-represented minority students. Thirty of these students were identified as having very low prior content in chemistry (approximately the lower third of content knowledge based on the overall student population in general chemistry at Johns Hopkins) and it was recommended that they either complete the studio course or include the problem-solving course. These students distributed themselves equally between the traditional lecture and laboratory course and the studio chemistry course. However, while students with low prior content represented 52% of the studio course, they represented only a small fraction of the lecture course. The remaining 36 students who were tested had a variety of scores, five students who were just above the recommended cut off for low content knowledge chose to take the studio course with their friends. When considering all students, the math SAT scores between the students enrolled in the studio course and in the traditional lecture was quite different. The average math SAT for students in the studio course was 686 (standard deviation: 54) compared to an average math SAT of 740 (standard deviation: 48) for the traditional lecture. Of note, the 25% percentile math SAT at Johns Hopkins for the Fall 2016 entering class was 710. 66% of the studio class had math SAT below 710 compared to the 22% of the lecture course. It has been found that persistence in STEM fields is generally under 20% for students who are in the bottom third of the math SAT for their school, regardless of their actual score (16).

Common Exam

Both the lecture course and the studio course included a total of four exams; three over the course of the semester and a cumulative final exam. The laboratory course also had two exams, whereas the laboratory concepts were incorporated into the examinations in the studio course. For the third exam, a common 70-point exam was prepared by one of the two lecture professors for their section with input from the studio professor. The lecture course took the 70-point exam during a 50-minute examination period, and the studio students took a 90-minute exam consisting of the common 70 points and an additional 30 points covering additional content due to different timing of the second exam. Due to the scheduling of different sections, only five of the fifteen students recommended for intervention were in the section of the lecture course that participated in the common exam. The results of the common exam are shown in Table 1. While the students in the studio course had a lower overall median than the students in the large lecture course, this is to be expected because most of them entered with significantly lower content knowledge and math SAT scores. While the sample numbers are small, what is notable that a full third of students in the studio course were able to obtain an above median score and a significant portion where able to obtain very high scores. This suggests that the studio format provided an avenue for students with low initial content knowledge to succeed in the course. 135

Table 1. Comparison of Students Selected for Extra Intervention on 70 Points of Identical Questions Recommended for intervention Lecture

Lecture

Studio

Median

56

50

51

Average

53.5

48.2

49.2

Range

24-70

40-53

18-69

% over 56/70

45%

0%

33%

Number

300

5

15

% over 60/70

27%

0%

20%

Validation of Consistency in Student Grades Every effort was made to have a standardized set of expectations, and consistent grading between the sections. Complicating the comparison between the two classes is that the studio course is a 4-credit lecture and lab combination course compared to a 3-credit lecture course and a 1-credit laboratory course. A combined lecture and lab grade for the traditional model students was calculated by combining 75% of their lecture grade and 25% of their lab grade. The common third exam provides an opportunity to calibrate the grading between the two classes. Regression analysis was used to correlate the third exam grade to either the final lecture grade or the combined grade. The assigned grade in the studio chemistry course was used for both analyses. As expected, the correlation between the 70-point third exam and overall grade was very high. A variable representing being enrolled in the studio course compared to the traditional lecture was added to the regression analysis, but it was highly insignificant (P=.95, .90) while the exam 3 grade was highly significant. The common exam was given in only one lecture section. However, students from both lecture sections are in the same lab course, which is a single course taught by one professor. The lecture grades are highly correlated with the lab grades (R2=.55, P