Understanding Research Methods: An Overview of the Essentials [Paperback ed.] 1884585647, 9781884585647

Citation preview

Fifth Edition

Understanding

Research

Methods mwiLU

imm

mooo i ioioq

A

m

nmujjl

UIOIOIIU

Tftll/w

Mildred

An Overview

L.

Patten

of the Essentials

Digitized by the Internet Archive in

2013

http://archive.org/details/understandingresOOpatt

Understanding

Research Methods An Overview of the Essentials Fifth Edition

Mildred L. Patten

Pyrczak Publishing P.O. Box 250430

'Pyrczak Publishing"

is

•

Glendale,

CA 91225

an imprint of Fred Pyrczak, Publisher,

A

California Corporation.

Although the author and publisher have made every effort to ensure the accuracy and completeness of information contained in this book, we assume no responsibility for errors, inaccuracies, omissions, or any inconsistency herein.

Any

slights

of people, places, or organizations are

unintentional.

This edition was written in collaboration with Randall R. Bruce. Project director:

Monica Lopez.

Editorial assistance provided

by Cheryl Alcorn, Karen M. Disner, Brenda Koplin, Erica Simmons,

and Sharon Young.

Cover design by Robert Kibler and Larry Nichols. Printed in the United States of America by Malloy, Inc.

© 2005, 2004, 2002, 2000,

1997 by Pyrczak Publishing. All rights reserved. No portion of this book may be reproduced or transmitted in any form or by any means without the prior written permission of the publisher.

Copyright

ISBN 1-884585-64-7

111

....

Contents Introduction to the Fifth Edition

Part

A

vii

Introduction to Research Methods

1

1

Introduction to Empirical Research

3

2.

Experimental

vs.

Nonexperimental Studies

5

3.

Experimental

vs.

Causal-Comparative Studies

7

4.

Types of Nonexperimental Research

5.

Variables in Nonexperimental Studies

1

6.

Variables in Experimental Studies

13

7.

Research Hypotheses, Purposes, and Questions

15

8.

Operational Definitions of Variables

17

9.

Quantitative vs. Qualitative Research:

9

I

10. Quantitative vs. Qualitative Research: II 1 1

Program Evaluation

Part

B

The Role of Theory

C

27

Research

29

Reasons for Reviewing Literature

3

15.

Locating Literature Electronically

33

16.

Organizing a Literature Review

35

17.

Writing a Literature Review

37 41

References

43

Sampling

21

Random and

Stratified

22. Other

Systematic Sampling

Methods of Sampling

5

53

Sample Size

A Closer Look at Sample Size

55

57

Measurement

59

25. Introduction to Validity 26. Judgmental Validity

6mkfil

"

27. Empirical Validity (Cfityttf*)

16CJL

rV

and

Its

Relationship to Validity

63

71

75

30.

Measures of Reliability

3

Norm- and Criterion-Referenced

1

61

67

28. Judgmental-Empirical Validity 29. Reliability

47 49

Random Sampling

23. Introduction to 24.

45

Biased and Unbiased Sampling

20. Simple

D

25

14.

19.

Part

in

Research

Reviewing Literature

18. Citing

Part

21

23

12. Ethical Considerations in 13.

19

Tests

77

32.

Measures of Optimum Performance

79

33.

Measures of Typical Performance

81 Continued ->

Hi

51

. .

Part

E

Part F

Experimental Design

83

34. True Experimental Designs

85

35. Threats to Internal Validity

87

36. Threats to External Validity

89

37. Pre- Experimental Designs

91

38. Quasi-Experimental Designs

93

Understanding

95

Statistics

97

39. Descriptive and Inferential Statistics

99

40. Introduction to the Null Hypothesis

41

Scales of Measurement

K10/£

101

42. Descriptions of Nominal Data

103

43. Introduction to the Chi Square Test

105

44.

A

Closer Look

at the

Chi Square Test

1

109

45. Shapes of Distributions

46.

The Mean, Median, and Mode

1 1

47.

The Mean and Standard Deviation

1

48.

The Median and

49.

The Pearson Correlation Coefficient

117

50.

The /Test

119

5

One-Way

Analysis of Variance

Two-Way

Analysis of Variance

1

52.

Interquartile

Range

Effect Size

1

125

127 129

54. Introduction to Effect Size (d)

of Effect Size (d)

and Correlation

133

57. Introduction to Meta-Analysis

135

58. Meta-Analysis

137

and Effect Size

59. Meta-Analysis: Strengths

H

131

(r)

56. Effect Size

Part

2

123

and Meta-Analysis

55. Interpretation

13

1 1

53. Practical Significance of Results

PartG

07

and Weaknesses

Qualitative Research

139 141

60.

Sampling

in Qualitative

Research:

I

143

61.

Sampling

in Qualitative

Research:

II

145

62.

Measurement

in Qualitative

63.

Measurement

in Qualitative Research: II

Research:

I

64. Quality Control in Qualitative Research 65.

Grounded Theory

for

Data Analysis

66. Consensual Qualitative Research: Analysis

147

149 151 1

53

155

Appendices A. Outline of Question Route for a Focus Group Study of Stress

157

IV

1

.

B. Excerpts from Literature Reviews

159

C. Electronic Databases for Locating Literature

163

D. Electronic Sources of Statistical Information

165

E. Other F.

G.

Methods of Determining

Reliability

A Closer Look at the Standard Deviation A Closer Look at Effect Size

1

7

173 175

Tables

Index

1

Table of Random Numbers

1

2.

Table of Recommended Sample Sizes

179

77

181

VI

Introduction to the Fifth Edition This book provides you with an overview of basic

pendent on welfare, which types of educational programs to fund, and how to adjust work en-

research methods.

vironments to improve employees' output and

The •

distinctive features of this

The

book

are:

satisfaction. If

maker in your methods to be

division of the material into short sections

instead of long chapters will help you take

small steps through this exciting technical

—

—but highly

of study. The long chapters

field

which often are not

literature

in

improvements made by

one topic builds directly on the previous one, the second one begins with a reminder of what you should have mastered in the previous a

smooth

transition

Technical jargon

pected to collect data on their clients.

from one

you will be making lifestyle decisions on research reported in the media. Should you take vitamin supplements? How should you dress for success on the job? Which make of automobile should you buy if your primary concern is safety? Answers based on research are often offered in newspapers, magazines, and television newscasts. As a result of studying research methods, you will become a knowledgeable, critical consumer of research. For instance, you will consider questions such as "Was the sample biased?" "Is the data analysis misleading?" and "Are the results All of

•

defined in plain English to

is

and numerous examples

abstract research concepts concrete. Stu-

dents' reactions in field tests attest to success in the effort to

•

make

The material on statistical

The

book comprehensible.

statistics is

presented

at the

It

with computational •

this

shows you how to interpret reports but does not bog you down

conceptual level.

new methods

based

the extent possible,

make

their clients, teachers

classroom, and social workers are ex-

in the

topic to the next. •

do simple but

to

are expected to experiment with

This helps you keep your eye on the big

make

be expected

will

psychologists are expected to track

clinical

picture and

topic.

important research on the job. For instance,

When

topic.

on a

Many of you

•

easily

digested. •

effective in sorting through the

conflicting claims often found in the research

other research methods books prompt students to take big gulps,

you hope to become a decisionfield, you must master research

details.

exercises at the end of the topics encour-

statistically significant?"

make sure you have mastered one topic before moving on to another. This is important because many topics are cuage you to pause to

mulative. is

Thorough mastery of an

•

Finally,

with ease. The

first

to read

and report on

solid

part of each ex-

You

will

you have a understanding of basic methods of re-

be more skilled

earlier topic

frequently a prerequisite for mastering a

later topic

you may need

published research in other classes.

doing

at

this

if

search.

your comprehension of factual material. The second part asks you to interpret and apply the material you have mastered. This ercise tests

will help

you

New to

This edition contains a

internalize the concepts as well

sider

major issues

in

you con-

planning a research pro-

ject.

64 describes

should you have an overview of research methods? Because...

Part

H

scribe

Leaders

in all fields are increasingly relying

part (Part H),

which

specific

techniques

qualitative

re-

searchers use to ensure the quality of their research.

Why •

new

expands the coverage on qualitative research. Specifically, Topics 60 and 61 describe sampling in qualitative research, Topics 62 and 63 describe methods used to collect qualitative data, and Topic

as stimulate classroom discussions. Finally, the third part provides questions that help

the Fifth Edition

concludes with Topics 65 and 66, which detwo frequently used general approaches to

the analysis of qualitative data.

on

the results of research in making important decisions, such as how to help those who are de-

vu

Acknowledgments _ _ _ .. ,. TT ~ Dr. Anne Hainer and Dr. Robert Morman, both .

.

--,..-.„ of California State ...

vided ,

.

this

,

many ,

book.

,

,

_

New

,

•

.

"

comments on .

,

c

,

.

^

..,

_° Z . the first draft or

material for this edition

was

.

A11

, ,

,.

..

.,

Jr

..

.

,

,

made important con... f . „ development orc this book. How.

,

.

ever, errors and omissions, , ., ri responsibility of the author,

T

re-

_ _, _ . ,„ , , ,,, viewed by Dr. Robert Rosenthal of Harvard University and the University of California, Riverside; Dr. ,

of California State University, Los

tnbutions to the

,

...

Oh

College. All these individuals

,

University, Los Angeles, pro-

.

helpful

VT

TT

Deborah M.

Angeles; and Dr. Richard Rasor of American River

,

•

.

c

Mildred L. Patten

Vlll

..

of course, remain the

,

Part A Introduction to Research Methods This part of the book defines what

view of the

characteristics

underlie

types of research

— such define they plan between theory and research — all

the variables

ship

we mean by

of the major approaches

empirical research and provides an overto this type

of research. Broad issues that

as the nature of research hypotheses,

to study, ethical considerations in research,

are also covered in this part.

how

researchers

and the

relation-

Topic

Introduction to Empirical Research

1

The empirical approach to knowledge is based on observations. We all use the empirical approach in everyday living. For instance, if a teacher observes students becoming restless during a certain lesson, he or she might say they "know" the lesson 1

is

boring.

are,

As

useful as everyday observations often

they can be misleading and are often misinter-

may have

preted. For example, the teacher

misin-

terpreted reasons for the students' restlessness.

It

might be the time and day, such as a warm Friday afternoon, that is the culprit and not the inherent interest of the lesson. Even if the lesson is boring to this teacher's students, the teacher

lesson

that the

when

it

might conclude

boring to students

is

might, in

fact,

be interesting

in general,

to other stu-

dents at other ability levels, with different back-

grounds, and so on.

When

researchers use the empirical approach,

be observed, which

how

often the case, they consider

is

sample

to obtain a

biased against those

ods of drawing unbiased samples are discussed

C of this book. When researchers

lect

among

plan

to

make

careful plan-

observations,

they want to observe, as well as

to

observe.

with an eye to selecting the most valid instru-

none

judged to be reasonably valid for develop new instruments. Then, of course, they need to decide when they will use ments).

If

is

their purpose, they

the instruments to obtain the

to

observe establishes the

most valid

results.

D

in detail in Part

of this

The observations researchers make may result in in the form of numbers, which are analyzed

data

statistically.

The question of why

to observe, they se-

interviews, and direct observation of behavior,

tests,

whom

how and when

how

available instruments such as objective

book.

is

in

Part

The key ning of why they want

doing so

not biased against

Such a sample would be who are less interested. Meth-

These issues are explored

to

is

the population as a whole.

they strive to avoid misleading results and poor interpretations.

that

any types of individuals or subgroups. For example, asking students to volunteer to take a mathematics lesson might result in a sample of students who are more interested in mathematics than the students in

Popular

scribed in Part

statistical

techniques are de-

F of this book. Note

that

some

scien-

observations are not reduced to numbers but

need for the study. Perhaps a better method for

tific

helping students acquire a certain mathematics

are expressed in words. For example, interview data

is

needed. After considering their

own

skill

experiences

may be

described in a narrative that points out

on the topic, researchers prepare a formal statement of their research purpose, such as "whether the use of handson manipulatives to teach Topic X will result in greater student achievement than a lecture and

themes and trends. The choice between the two approaches is described in Topics 9 and 10 as well as Part H of this book. One of the most fundamental distinctions in scientific research is whether research is experimental

workbook approach." They might also arrive at a hypothesis, which is a statement of what they ex-

or

pect the results to show. For example, they might

as treating

and reviewing related

literature

who

nonexperimental.

In

experimental

research,

treatments are given for the research purpose, such

some

students with manipulatives and

use manipulatives will

others with a lecture/workbook approach in order to

have higher scores than those who are exposed to the lecture/workbook approach. The question of why is explored throughout Parts A and B of this

determine which treatment causes greater achieve2 ment. Of course, we are not always interested in

book. decide whether to observe an entire population

want to know whether teachers believe they need more training in the use of manipulatives for teaching mathematics. For this particular research pur-

fifth-grade students in a school district)

pose, a researcher should ask teachers about their

hypothesize that those

When first

(such as

researchers plan

all

whom

to observe, they

or just a sample of the population. If a sample

is

to

cause-and-effect questions. For example,

needs;

it is

we might

not necessary to train or treat the teach-

ers to obtain the research results. Thus, nonexperi-

mental research 1

Examples of other approaches are (1) deduction, such as when we deduce a proof in mathematics based on certain assumptions and definitions, and (2) reliance on authority, such as relying on a dictator's pronounce-

The

ments as a source of knowledge.

book.

2

distinction

needed for this research purpose. between experimental and nonex-

is

The design of experiments

is

discussed in Part

E of this

perimental research ics

is

explored

in the

next two top-

of this book.

Exercise on Topic 1

The empirical approach

2. Is the

to

knowledge

1

based on what?

is

empirical approach used in everyday living? the question of why establish?

3.

What does

4.

How

5.

According

6.

What do

7.

Are the

8.

Are treatments given

the term hypothesis defined in this topic?

is

to the topic, are

researchers do

results

of all

samples often observed?

when

they plan

how to

observe?

scientific studies expressed as

numbers?

for the research purpose in experimental research or in nonexperimental research?

you asked students for their opinions on switching from the semester system would you be conducting experimental research or nonexperimental research?

9. If

to the quarter system,

Questions for Discussion 10. Briefly describe a

time

when you were misled by everyday

sion based on everyday observation that

1 1

observation

(i.e.,

when you reached

a conclu-

decided was an incorrect conclusion).

probably encountered conflicting research reported in the mass media. For example, one study might indicate that X increases blood pressure while another study indicates that X does not increase it. Speculate on the reasons why various researchers might obtain different results when studying the same problem.

Name still

a general

Who Are Planning Research problem area

in

which you might conduct research. At

be broad, such as "social phobia," or

tion in the treatment

search and

13.

later

You have

For Students 12.

you

make

it

may be

of social phobia." Note

that

this point,

your problem area may

narrow, such as "effectiveness of behavior modifica-

you may want

to

name

several

problem areas

for re-

a final selection at a later time.

Have you already made observations that observations

may be

direct,

in your problem area(s)? If so, briefly describe them. (Keep in mind such as observing aggressive behavior on a playground, or indirect, such

as asking adolescents for self-reports

on

their alcohol

consumption.)

you answered yes to Question 13, did you make informal observations or did you plan why, whom, how, and when to observe in advance of making the observations? Explain.

14. If

Topic 2 Experimental vs. Nonexperimental Studies

experiments, researchers give treatments and if they cause changes in behavior. A classic simple experiment is one in which we form

that if a researcher administers treatments or ar-

random and give each group a differTo form two groups at random, we can put the names of the available participants on slips of paper, mix them thoroughly, and pull some names for each group. Notice that random assignment gives each participant an equal chance of be-

are

In

observe to see

two groups

at

ent treatment.

ranges for their administration, the study

is

called

an experiment whether or not groups of participants

formed

at

random.

Some experiments

are conducted with only one group of participants, as illustrated in Example 3, in which the extra verbal praise is the treatment.

1

The

ing in either group.

Below

are three

Example

3

students in one classroom were observed for an

hour each day for a week

examples of experiments.

in order to

count the num-

ber of inappropriate out-of-seat behaviors. During the

next week, the teacher provided extra verbal praise

Example

1

were divided into two groups at random. One group received math instruction via a correspondence course on the Internet. The other group was given instruction on the same math skills using a traditional textbook and workbook approach. The purpose was to see if instruction via the Internet was Fifty students

more

when

providing the extra verbal praise. The results showed less

inappropriate

out-of-seat

behavior during the

second week of the experiment.

In

effective than traditional instruction.

were in their seats at appropriate During the third week, the teacher stopped

students

times.

nonexperimental studies, researchers do not

give treatments. Rather, they observe participants in In

Example

1

,

the group receiving the

of instruction via the Internet

is

new

type

referred to as the

experimental group while the group receiving ditional instruction is called the control

When

true experiment. Not

all

2

experiment

is

who

will

called a

experiments are true ex-

periments, as illustrated by

poll in

Example

Nonexperimental studies come

which

volunteered to take a

direction.

She

new drug under

her

identified an additional 100 nonvolun-

same diagnosis and similar demographbackground characteristics such as age and

teers with the ics (i.e.,

gender) to serve as controls. The study was con-

ducted to investigate the effectiveness of the

drug

new

in treating depression.

experiment was conducted by comparing the volunteers who were given the new drug with a group of nonvolunteers who were not In

Example

given the

new

2, the

drug. This study

is

an experiment

even though random assignment was not used. Note

forms,

more detail in Topic 4. At however, you should be able to distinstudies

and ex-

administered for experimental purposes.

you cannot distinguish between nonexperimental and experimental studies on the basis Note

Other methods for drawing random samples are

dis-

in Part C of this book. Types of experiments are explored more fully in Part E, where you will learn the advantages of true experiments.

that

of the type of instrument (i.e., measuring tool) used. Instruments such as paper-and-pencil tests, interview schedules, and personality scales are used in both types of studies. The act of measurement is usually not considered to be a treatment. In fact, researchers try to measure in such a

way

that the act

of measuring does not affect or change the participants. This is true in both experimental and nonexperimental studies.

By now, you may have inferred that the purpose of an experiment is to explore cause-and-effect rela(i.e.,

treatments are given to see

learn

how

nonexperimental studies are

times used for

this

purpose.

how

they

you will also some-

affect the participants). In the next topic,

cussed 2

many

are explored in

tionships 1

in

periments by determining whether treatments were

psychiatrist identified 100 clinically depressed cli-

who

a sur-

they exist without experimental intervention.

this point,

2.

Example 2 ents

is

which participants are interviewed,

guish between nonexperimental

A

most

the

termine their attitudes, beliefs, and behaviors as

experimental group and

in the control group), the

One of

random

who be

intervention.

types of nonexperimental studies

hat to determine

names out of a

in the

common

questioned, or otherwise observed in order to de-

the participants are divided at

be

out experimental

vey or

group.

(such as drawing will

tra-

order to describe them as they naturally exist with-

Exercise on Topic 2 1

Are treatments given

2. In

4.

6.

it

is

A

was given verbal praise Which group is the

special treatment.

necessary to have

What

5. Is

nonexperimental studies?

an experiment, Group

was given no 3. Is

in

at least

for being

on time

for

appointments while Group

B

control group?

two groups of participants

in order to

conduct an experiment?

the purpose of a nonexperimental study?

a survey an experiment?

Does knowing

that a multiple-choice test

was used

in a study help

you determine whether the study was

experimental or nonexperimental?

7.

What

8.

A

is

social

the

major purpose of an experiment?

worker polled

clients to determine their satisfaction with the services provided. Is this an "ex-

perimental study" or "nonexperimental study"?

9.

A teacher tried three methods of teaching handwriting by using different methods with different students. Is this

an "experimental study" or "nonexperimental study"?

Questions for Discussion 10.

Suppose you read

that an outbreak

contaminated chicken served

in a

of intestinal disorders occurred

popular restaurant.

"experimental" or "nonexperimental"?

1 1

Is

it

in a

town and the source was traced to was

likely the study that identified the source

Why?

Have you ever conducted an informal experiment by giving a treatment to a person or a group and then observing the effects? If so, briefly describe it. Would you have obtained better information by including a control group? Explain.

12.

know whether reading to preschool children has a positive effect on subsequent Do you think it would be better to conduct an "experimental study" or "nonexWhy?

Suppose you wanted

to

reading achievement. perimental study"?

For Students 13.

At

this point,

Who Are Planning Research

do you anticipate using an "experimental" or "nonexperimental" approach

in

your

re-

search? If it will be experimental, what treatments do you plan to administer?

you plan to conduct experimental research, have you already conducted an informal experiment on your topic? If so, did you use a control group? Was it a true experiment? How will the experiment you are planning differ from the informal experiment you already conducted?

14. If

Topic

Experimental vs. Causal-Comparative Studies

3

As you know from Topic 2, an experiment is a study in which treatments are given in order to ob-

while most of those in the control group reside

in

(such as requiring them to

Because urban areas tend to have more smog than rural areas, smog might be the cause, and smoking might be coincidental. These types of problems would not arise in an experiment in which participants are divided at random to form two groups: one of which is made to smoke and the other forbidden to smoke. They would not exist because the random assignment would produce two groups that are equally likely to experience stress and equally likely to live in either rural or urban areas and, in fact, be about equal in terms of all 2 other potential causes of cancer. The example of smoking and lung cancer illustrates a specific type of nonexperimental study known as a causal-comparative study (sometimes

rettes a

called an ex post facto study).

serve their effects.

we

"Does

ask,

stimulus) given

When we

conduct experiments,

the treatment

(i.e.,

the

input or

by the researcher cause changes

participants' behavior

changes

(i.e.,

in the

in

output or

response)?"

When

researchers want to investigate cause-and-

effect relationships, they usually prefer tal

experimen-

over nonexperimental studies. However, some-

times

it

not possible to conduct an experiment for

is

physical, ethical, legal, or financial reasons.

ample

the effects of

is

smoking on

health.

It

be unethical (because of potential harm

some

participants) to treat

participants with

An

ex-

would to

the

smoke

smoke a pack of cigaday for 15 years) in order to observe the effects in comparison with a nonsmoking control group (which is forbidden to smoke for 15 years). Clearly, for this research problem, we cannot conduct an experiment. Notice that even if it were ethical to conduct such an experiment, it might not be practical because we probably would not want to wait 15 years to determine the answer to such an it

is

who

we must

information derived from nonexperimental

we

can identify both people currently have lung cancer and a control group

studies.

For example,

with similar demographics

(i.e.,

background char-

such as socioeconomic status) and describe the differences between the two groups in terms of previous lifestyle characteristics that might

acteristics

affect health

such as

scription drug use,

3

The

essential char-

of this type of nonexperimental study are observe and describe some current condition

acteristics

(1)

we

(such as lung cancer) and (2) we look to the past to try to identify the possible cause(s) of the condition.

Notice that researchers do not give treatments

in

causal-comparative studies. Instead, they only deobservations.

scribe

Hence, they are conducting

Although the causal-comparative method has

impossible or impractical to conduct

an experiment to answer a causal question, settle for

1

nonexperimental studies.

important question.

When

rural areas.

diet, exercise,

illicit

more

potential

method,

it is

pitfalls

than

the

experimental

when when it is

often the best researchers can do

attempting to explore causality. Note that

used properly, and the comparison groups are selected carefully, the causal-comparative method is a powerful scientific tool that provides data on

many

important issues in

all

the sciences.

smoking, pre-

substance abuse, and so

on.

A

finding that smoking differentiates between two groups that have similar demographics suggests that smoking is a possible cause of lung cancer. However, there are several dangers in this interpretation. First, smoking and cancer might have a the

common

cause. For example, perhaps stress causes

cancer and also causes people to smoke

excessively.

banning smoking will not prevent cancer, only reducing stress will. Another danger is that the researcher may have failed to identify control participants who were properly matched with If this is the case,

who have

lung cancer. For instance, perhaps most of those with lung cancer reside in urban areas those

Because of the laws of probability, the larger the sammore likely that two groups formed at random will be equal in terms of their characteristics. Sample size is covered in Topics 23 and 24. 2 The relationship between smoking and health has been 1

ple, the

examined ies.

On

in

many hundreds of causal-comparative

this basis,

almost

all

stud-

experts agree that alternative

However, the results of effects of smoking on health were disputed because the studies were not true experiments (i.e., did not have random assignment to interpretations are without merit.

some of

the early studies

on the

groups). 3

Other types of nonexperimental studies are covered

the next topic.

in

Exercise on Topic 1

According

when 2.

We

3. Is

to the topic,

3

do "experimental" or "causal-comparative" studies have more potential

pitfalls

trying to identify cause-and-effect relationships?

look to the past for a cause in which type of study?

causal-comparative research a type of experiment?

4.

Are treatments given by researchers

in

5.

Random assignment

used in which type of study?

6.

How

7.

A researcher compared the health of adolescents who received free

is

the term

to treatments is

demographics defined

causal-comparative studies?

in this topic?

school years with the health of a comparable group of children

who

lunches during their elementary did not receive free lunches.

The

purpose was to determine the effects of free lunch on health. Did the researcher conduct an "experimental"

8.

or a "causal-comparative" study?

A researcher divided patients with Upon

their release, the researcher

diabetes

who were

being released from the hospital into two groups.

provided normal counseling for individuals with diabetes to one group

while providing the other group with extended counseling. The purpose was to determine the effects of the ter

9.

two types of counseling on patients' compliance with physicians' directions during the first month hospitalization. Did the researcher conduct an "experimental" or a "causal-comparative" study?

What

is

another

name

for the

af-

term causal-comparative study]

Questions for Discussion 10. If you

wanted

to investigate the

causes of child abuse, would you use the "experimental" or the "causal-

comparative" method? Explain.

1 1

Suppose you read that a causal-comparative study indicates that those who take vitamins A and E tend to be less overweight than the general population. Are there any possible dangers in the interpretation that the vitamins cause people to maintain a healthy weight?

For Students 12. If you will

Who Are Planning Research

be conducting a nonexperimental study, will it be causal-comparative (i.e., looking to the past of some current condition)? If yes, briefly explain why you chose this method of research

for the causes

instead of the experimental method.

Topic 4 Types of Nonexperimental Research

As you know from Topics 2 and 3, researchers do not give treatments to participants in nonexperimental studies. Rather, they observe (i.e., measure) in

order to describe the participants without trying

to

change them.

longitudinal

ing

week

ple of infants each

and because people vary or

learned in the previous topic about causal-

comparative research

which researchers look to the cause(s) of a current condition. It is

causality, but

when it

is

in

researchers are interested in

not possible to conduct an ex-

periment.

Another type of nonexperimental research is the survey or poll. The purpose of surveys is to describe

the

attitudes,

population.

re-

In correlational research, researchers are inter-

briefly described here.

used primarily

a

for a year to trace their

more

the past for

instance,

development.

Nonexperimental studies take many forms because they serve many purposes. Some of the more common types of nonexperimental studies are

You

For

research.

searcher could measure the visual acuity of a sam-

We

beliefs,

and behaviors of a

draw a sample of

make

study the sample, and then

a population,

inferences to the

ested in the degree of relationship

a college admissions test and

they are variables. in

among two

or

quantitative variables. For instance, scores on

2

which he or she

GPAs

differ

are quantitative

on both of them,

If a researcher conducts a study is

asking, "Did those with high

GPAs?" and "Did those with low admissions scores tend to earn low GPAs?" the researcher is asking a correlational question. To the extent that the answer to the question is true, we can say the test works (i.e., has va3 lidity for predicting GPAs). Experiments and all the types of research mentioned so far in this topic belong to the class of research called quantitative. A distinctive feature of admissions scores tend to earn high

population from the sample data. For example, a

quantitative research

researcher could survey a sample of people receiv-

that are easy to quantify so that they

ing food stamps to determine what types of food

lyzed with

they purchase with the stamps. The results obtained

tudes toward Asian American immigrants, a quanti-

from studying the sample can be generalized

to the

tative

population (assuming that a good sample has been

count

drawn).

1

Note

that if

instead, interview all

decide not to sample but,

everyone

in the

population

(i.e.,

people receiving food stamps), the study would

be called a census. all

we

members of a

A

census

is

that researchers gather data

can be ana-

For instance, to measure

researcher might use a questionnaire

how many

atti-

and

times respondents answer "yes" to

statements about Asian Americans and then calcu-

percentage

late the

who answered

"yes" to each

statement. In contrast, qualitative researchers gather data

a count (or study) of

(such as responses to open-ended interview ques-

population.

While surveys usually include hundreds or thousands of participants, a case study usually involves only one. For example,

statistics.

is

some important

theories in

on attitudes toward Asian Americans) that must be analyzed using informed judgment to identify major and minor themes expressed by particitions

Most published

qualitative research

psychology were developed based on intensive one-on-one case studies of individuals. In a case study, the emphasis is on obtaining thorough knowledge of an individual, sometimes obtained over a long period of time. Researchers do not confine themselves to asking a limited number of questions on a one-shot basis as they would do in a

pants.

survey.

of this book describes qualitative research methods in detail. In historical research, information is examined in order to understand the past. Note that good his-

clinical

When

researchers repeatedly measure

traits

of

trace developmental trends,

Characteristics of

we

say

good samples

search are explored in detail in Part

we

are conduct-

for quantitative re-

C

of

this

book. Con-

siderations in sampling for qualitative research are de-

scribed in Topics 60 and 61 in Part

H

of this book.

is

a core

terviewers

col-

interviews in which

of questions from which the indeviate as needed to obtain in-depth

list

may

information.

Other differences between qualitative and quanresearch are explored in Topics 9 and 10.

titative

Part

the participants over a period of time in order to

1

lected using semi-structured

there

is

2 3

H

Types of variables are described Validity

is

tional studies ficient,

which

D

Topics 5 and 6. of this book. Correla-

in

explored

in Part

employ a

statistic called a

is

described

in

Topic 49.

correlation coef-

matter of developing a

torical research is not just a

chronological

of so-called

list

facts

and

are collected. Historical researchers

Rather, it is an attempt to understand the dynamics of human history. As such, it is driven by theories

and hypotheses. In other words, by reviewing historical evidence, researchers are able to develop theories that terns.

may

to hypotheses,

methods

use quali-

(e.g.,

ments). Historical research history departments and

explain historical events and pat-

These theories lead

may

examining historical documents using insight and judgment to identify themes) or qualitative methods (e.g., counting certain types of statements made in historical docutative

dates.

which are

this

is

is

typically taught in

not considered further in

book.

evaluated in terms of additional historical data that

Exercise on Topic 4 1.

Suppose a researcher administered an intelligence test to young children each year for five years to study changes in intelligence over time. The researcher was conducting what type of study?

2. Is

the study in Question

3. If a

1

in

order

experimental?

researcher conducts a poll to estimate public support for free child care for mothers on welfare, the

researcher

is

conducting what type of nonexperimental study?

4.

An

5.

According

6.

Most published

between vocabulary scores and reading comprehension scores. The researcher was conducting what type of nonexperimental study? investigator determined the degree of relationship

to this topic,

what

is

a distinctive feature of quantitative research?

qualitative research

is

collected using what?

Questions for Discussion 7.

Name

8.

Think of a survey

a topic in your field of study that you might explore with a nonexperimental study. Which type of nonexperimental study would be most appropriate for your topic?

which you were asked to serve as a participant. (You may have been sent a quesconsumer satisfaction survey, or you may have been contacted in person or by phone.) Did you cooperate and respond? Why? Why not? in

tionnaire in the mail, such as a

9.

10.

Name two

quantitative variables that might be studied using correlational research.

Suppose someone prepared a list of educational events and their dates of occurrence Would the list be an example of "good" historical research? Explain.

For Students 11. If

you

will

in the last

100 years.

Who Are Planning Research

be conducting a nonexperimental study, which type will

choice.

10

it

be? Explain the basis for your

Topic

A

variable

more

Variables

5

a trait or characteristic with

Nonexperimental Studies

in

two or

or her gender and "names" the preferred candidate.

categories. Participants vary in terms of the

Other variables have quantitative categories that describe the amount of a characteristic. Consider

categories to

is

which they belong. Example

1

has two

variables.

Example

Example

Example

1

A

sample of registered voters was surveyed. Each voter was asked first to name his or her gender (male or female) and then to

name

the candidate for

whom

man GPAs predicting

preferences.

1

who

dates).

careful not to confuse a variable with

For instance, "male"

categories.

is

One

Two

is

how

one of the two "Male" is

to visualize

Categories =>

Male

That

is,

long to one

earns an

able.

we would

not allow a participant to

such a survey,

candidates

we

A

a category.

student

to the cate-

(i.e.,

two categories)

variable (stimulus or input)

causes changes in the

dependent variable (response or output).

ask,

Some comes

for the

researchers

first

refer

(whether or not

to it

also try to define variables in such a

way

later as

To be exhaus-

any variable that presumed to be a the one that comes

is

cause) as "independent" and to

political office.

that the categories are exhaustive.

"dependent." For instance,

SAT

scores (the

predictor variable) are usually determined before

must have a category for each we would want to add a fourth choice of "undecided" for respondents who have not yet chosen among the three candidates. Note that "undecided" is also a mutually exclusive category because respondents would not be allowed to choose it as well as name one of the candidates. Notice that both variables in Example 1 are "naming" variables (more properly called "categorical" variables).' Each participant "names" his tive,

is

of 550 belongs

Remember:

The independent

category. For instance,

candidate do you plan to

We

SAT score

pendent variable, and the response or outcome (such as lung cancer) is called the dependent vari-

Female

mutu"For which one vote?" When conducting

same

for

each respondent to the survey will be-

—and only one—

exclusive because

name two

is

pendent. For example, when we conduct a causalcomparative study (see Topic 3 to review), the presumed cause (such as smoking) is called the inde-

it:

the categories for "preferred candidate" are ally

SAT

Variables in nonexperimental studies are some-

All variables have mutually exclusive categories.

valid the

times classified as being either independent or de-

GENDER

Variable =>

how

gory called "550." Likewise, a student who has a GPA of 2.78 belongs to the category called "2.78."

its

categories of the variable called "gender."

not a variable. Here

determine

words, each possible score

three possible candi-

(i.e.,

to

GPAs.

Both the College Board's SAT scores (ranging from 200 to 800 per subtest) and GPAs (usually ranging from 0.00 to 4.00) are quantitative. In other

The two variables in Example are (1) gender, with two categories and (2) the preferred candidate, with three categories

2

The college admissions scores on the Scholastic Aptitude Test (SAT) were compared with students' fresh-

he or she planned to vote (Doe, Jones, or Smith). The purpose was to explore gender differences in voting

Be

2.

a variable

GPAs. Thus, some reSA T the independent vari-

students earn their college

respondent. For instance,

searchers

would

able and the

common

call the

GPA

the dependent variable.

to call the predictor variable the

It is

also

predictor

and the outcome variable (such as GPA) the criterion. The term "criterion" means "standard." Hence, GPA is the standard by which the SAT is often judged. Ideally, high scores on the SA T should be associated with high GPAs while low SAT scores should be associated with low GPAs. To the extent that this is true, the SAT is judged to be adequate. To the extent that it is not true, the SA T is judged to

Naming data are obtained from the nominal scale of measurement. Scales of measurement are discussed in Topic 41.

be deficient. Procedures for determining the adequacy of a test are described in Part D of this book.

11

Exercise on Topic 1

Adults

who were

5

taking a course to learn English as a second language were asked to

of birth and their number of years of formal education.

In this

how many

example,

name

their country

variables were being

studied?

Question

1,

which variable

is

a categorical variable?

3.

In Question

1,

which variable

is

quantitative?

4.

A sample of adults was asked their level

2. In

States

is

doing a good job

of agreement with the statement, "The President of the United

in foreign relations."

"agree," "disagree," or "strongly disagree."

5.

What

6.

A researcher looked

is

we

of social unrest by examining economic variables such as poverty, social unrest an "independent" or "dependent" variable?

Is

administer a basic math test to middle school children to see is

basic

if the

math

test scores predict

the criterion variable?

Suppose you ask participants to identify their age group using these categories: "under 21," "21-39," "40-55," and "56+." Are these exhaustive categories?

9. In

10.

respond either "strongly agree,"

to

were being studied?

for the causes

grades in high school algebra, what

8.

variables

meant by mutually exclusive categories?

income, and so on. 7. If

They were permitted

How many

Question

What

is

the

8,

how many

variables are being studied?

minimum number of categories

that

each variable must have?

Questions for Discussion 1 1

Suppose you want to measure income on a self-report questionnaire on which each participant will check off his or her income category. Name the categories you would use. Are they exhaustive and mutually exclusive? Explain.

12.

Name

a quantitative variable of interest to

you and name

its

categories.

Are the categories mutually ex-

clusive and exhaustive? Explain.

For Students 13. If you will

Who Are Planning Research

be conducting a nonexperimental study, name the major variables you will be studying. For

each, indicate whether the categories will be quantitative.

12

.

Topic 6 Variables in Experimental Studies

All experiments have at least one independent

the

of experiments

is

to estimate the extent to

which

independent variables cause changes in dependent variables.

As you know from dependent variable

Note

that

is

manipulate,"

we mean

physical

thing

of combinations of variables on causality. 2, which has two independent variables (i.e., child care and transportation money).

experiments, researchers physically

in

the

to

By

effects

Consider Example

"physically

the researcher does some-

Examples of

participants.

Example 2

physical manipulation are (1) giving a new drug to some participants while giving a placebo to others, (2) providing

some

Voluntary, free job training was offered to

all mothon welfare in a small city. Four groups of the mothers were formed at random to explore the effects of these two independent variables: (1) providing or not providing free child care while in training and (2)

ers

students with computers while

denying computers to others, and (3) using group counseling with some clients while using individual counseling with others. Thus, to physically manipulate

ers

means

providing or not providing transportation

to physically administer treatments.

Note that in nonexperimental studies, researchdo not physically manipulate independent vari-

ables. Instead, they

we

smoked

random to one of shown here:

the four treatment conditions

at

GROUP

who have nonexperimental studies we

Example

Example

On

—

and transportation

3

no child care and

money

transportation

money

GROUP 4

GROUP 2

no

child care

child care

no transportation

and no transportation

money

money

and

1

was predicted that those in Group 1 (the group that was given both child care and transportation money) would have the highest participation rates, those in Group 2 would have the next highest, those in Group 3 would have the next highest, and those in Group 4 would have the lowest. It

1

alternative weeks, a disruptive student

extra praise for being in his seat

The purpose of

the study

will increase the

is

when

is

given

appropriate.

to see if the extra praise

amount of appropriate

in-seat behav-

ior.

In

GROUP

1

child care

do not provide participants with cigarettes, nor do we expose them to smoke. (See Topic 3 to review other differences between experimental and nonexperimental studies of causation.) In a simple experiment, there is only one independent variable and only one dependent variable, as in

to

site.

observe independent variables

observe the health of people

cigarettes in

money

Each group was assigned

get to the job-training

as they occur (or have occurred) naturally. For ex-

ample,

toward school during the

attitude

ing than those with only one independent variable because such experiments give us insights into the

the previous section, an in-

a stimulus or input variable.

manipulate independent variables.

student's

weeks the extra praise is given. If we did this, we would have three dependent variables. Many experiments also have more than one independent variable. Often, these are more interest-

variable and one dependent variable. The purpose

Example

1

,

the physical manipulation

ing or not giving extra praise,

which

is

is

Notice that in Example

giv-

how

the inde-

is,

money is, and (3) the effectiveness of both child care and transportation money in combination. Thus, we get more information by looking at two independent variables

behavior.

more than one depend-

improvement

in

one study than by

each independent variable in a separate experiment (in which case, we could determine only

ent variable. For instance, in Example 1, we could observe to see not only if the treatment causes (1) more in-seat behavior, but also (2) improvement in

the student's achievement and (3)

we can determine (1) (2) how effective trans-

2,

portation

pendent variable (i.e., the stimulus or input variable). The dependent variable (i.e., response or outcome variable) is changes in the student's in-seat Often, experiments have

effective child care

in

13

looking

at

points

and

1

2).

Exercise on Topic 6 1

All experiments have at least

2. In

an experiment, what

is

the

how many dependent name of a

variables?

stimulus or input variable?

3.

What does physically manipulate mean

4.

Are dependent variables physically manipulated?

5.

Can an experiment have more than one independent

6.

Every other customer entering a shoe store was given a different coupon. One coupon offered a second pair of shoes for 50% off. The other coupon offered to reduce the total price by 25% if two pairs of shoes

in

an experimental context?

variable?

were purchased. The purpose was to determine which coupon was more effective buy two pairs of shoes. In this experiment, what is the independent variable?

what

In Question 6,

8.

A teacher showed an educational handout on the same material in

people to

the dependent variable?

7.

is

in getting

film on daily nutritional needs to one group of students and gave a

to another group.

The purpose was

to

determine which was more effective

increasing students' knowledge of daily nutritional needs. In this experiment, what

is

the dependent

variable?

9. In

Question

8,

what

is

the independent variable?

Questions for Discussion 10.

Name

a variable that

would be easy for you to physically manipulate by your manipulation.

in

an experiment. Then, name a

variable that might be affected

11.

Name

a variable that you would not be willing to physically manipulate for ethical or legal reasons.

For Students 12. If you will

Who Are Planning Research

be conducting an experiment, name the independent and dependent variables you will be

studying.

14

Topic 7 Research Hypotheses, Purposes, and Questions A research hypothesis is a prediction of the outcome of a study. The prediction may be based on an educated guess or a formal theory. Example 1 is

search purpose that corresponds to the nondirec-

Example

tional hypothesis in

Example 4

a hypothesis for a nonexperimental study.

The research purpose

Example ter

is

to explore the differences in

child-rearing practices between Tribe

1

hypothesized that first-grade

It is

3.

girls will

show

A research

reading comprehension than first-grade boys.

question

may

also be substituted for

Example

a nondirectional hypothesis.

Example

In

1

,

the author

is

predicting that he or

she will find higher comprehension than boys.

To

test the prediction, a

A and Tribe B.

bet-

among

5

shows a

research question that corresponds to the nondirectional hypothesis in

girls

nonexperimental

purpose

in

Example

Example

and the research

3

4.

study would be appropriate because nothing in the

Example 5

hypothesis suggests that treatments will be given.

A

The research question

simple research hypothesis predicts a relation-

ship between

two

variables.

is,

"How do

the child-rearing

A and Tribe B differ?"

practices in Tribe

From your study of

Topics 5 and

6, it should be clear that the two variExample 1 are (1) gender and (2) reading comprehension. The hypothesis states that reading

When

ables in

comprehension

Example 2

is is

research,

using a research question as the basis for

we

usually should be careful not to state

related to gender.

as a question that can be

a hypothesis for an experimental

"yes" or "no," as

is

done

it

answered with a simple

in

Example

6.

study.

Example 6 Example

The question

2

Tribe

It is hypothesized that children who are shown a video with mild violence will be more aggressive on

the playground than those

who

are

shown a

A

is,

and Tribe

"Do

B

Example 6 merely

similar

video without violence.

is

the child-rearing practices in

differ?"

asks,

"Do

they differ?" This

not a very interesting research question because

it

implies that the results of the research will be only a In

Example

lence (mild vs.

variable

is

2, the

independent variable

is

simple "yes" or "no." Example 5

vio-

cause

no violence), and the dependent

The hypotheses

in

Examples

1

we

predict

have more of some

which group

will

a research purpose,

be higher or

It is

attribute.

—

all

is

are acceptable

—not

should state a directional hypothesis

a re-

search purpose or question are suitable substitutes

from those of Tribe B.

for a nondirectional hypothesis.

The author of Example

3 is saying that there will

is

perfectly acceptable

inappropriate

It is

one as a substitute for a directional hypothe-

sis.

be a difference but does not predict the direction of

If

when

you have read research

reports,

you may have

encountered references to another type of hypothesis: the null hypothesis. This is a statistical hy-

no basis for making an educated guess as to the outcome of a study. Instead of a nondirectional hypothesis, we might state a research purpose. Example 4 shows a rethere

and a research question

search purpose or question. In other words, a re-

to use

the difference. This

a question

community. Of course, when we are willing and able to predict the outcome of a study,

3

are different

—

be complex and,

in the scientific

hypothesized that the child-rearing practices of

A

they differ?"

purely a matter of personal taste

we

Tribe

"How do

thus,

Sometimes we have a nondirectional hypothesis. Consider Example 3.

Example

asks,

more interesting and informative. The choice between a nondirectional hypothesis,

and 2 are exam-

ples of directional hypotheses. In a directional hypothesis,

it

that implies that the results will

aggressiveness on the playground.

superior be-

is

is

pothesis,

which needs

to

be considered when ana-

lyzing results obtained from samples in quantitative

research studies.

15

The

null hypothesis

need not be

considered

at

However, Part F of this book.

this point.

plored in detail in

it

will

be ex-

Exercise on Topic 7 1

Which type of statement

(hypothesis, purpose, or question) predicts the

hypothesized that college students

2. "It is

do not have firm career goals." 3.

who have

firm career goals achieve higher

GPAs

who

than those

a "directional" or "nondirectional" hypothesis?

Would an "experimental" or "nonexperimental"

study be better for testing the hypothesis in Question 2?

hypothesized that children of immigrants and children of native-born citizens will differ in their

4. "It is

attitudes

5.

Is this

outcome of a study?

toward school."

"The goal of this study

is

Is this

to

a "directional" or "nondirectional" hypothesis?

examine college students'

attitudes

toward religion."

Is this

statement a "hy-

pothesis" or "purpose"?

6.

"Are children of alcoholics different from children of nonalcoholics research question stated appropriately?

7.

When we

are willing to predict the

Why? Why

in their social

adjustment?"

Is this

not?

outcome of a study, should we

state a "directional"

or "nondirec-

tional" hypothesis?

8.

What

are the

two

alternatives to stating a nondirectional hypothesis?

Questions for Discussion 9.

Restate this hypothesis as a research purpose: "It tion lar

1

hypothesized that there their

is

a difference in job satisfac-

job performance and those

who

receive irregu-

feedback."

10. Is the

1

is

between those who receive regular feedback on

hypothesis in Question 9 "directional" or "nondirectional"? Explain.

Could an experiment be conducted

12. Restate this

those

hypothesis as a research question:

who do

"It is

Question 9? Explain.

hypothesized that those

who

exercise regularly and

not exercise regularly will differ in other behaviors that affect health."

For Students 13. State a

to test the hypothesis in

Who Are Planning Research

research hypothesis, purpose, or question for the research

you

are planning. (Note:

You may

more than one of each.) 14. If

you

stated a hypothesis in response to Question 13,

16

is it

"directional" or "nondirectional"?

have

'

Topic 8 Operational Definitions of Variables

Dictionaries provide us with conceptual defini-

of adequacy, and has been met.

For example, in research on speech communication, we might be interested in students' ability to recite, which has a dictionary

tions of variables.

definition along these lines:

we may

we

Example

Example 2

of our research to other individuals.

Positive self-esteem

when asked

to con-

on the

aids

we

definition

is

will

As we plan

ability to recite.

soon realize

not adequate because

it

The

will

take in order to identify the variable. Redefining a variable in terms of physical steps tionalizing a variable. variable,

we

Example

1

When we

is

first

operationalize a

attempt

"Do you

good about

feel

at creating

definition in

Example 2

we know what

is

reasonably opera-

specific question to

ask and what response to listen

for.

quite narrow. For example,

does not tap

esteem

called opera-

it

However,

it is

self-

dimensions of

clients' lives,

such as self-esteem in the workplace,

in social set-

in the various

and so on. Thus, Example 2 illustrates that a definition can be operational without being adequate in other respects, such as being fully multidi-

tings,

are creating an operational definition.

shows the

defined as answering "yes"

the question,

because

tional

does not indi-

we

is

2.

the

that a conceptual

cate the precise concrete or physical steps

not

yourself?"

duct an experiment on the effectiveness of two research,

is

it

could operationalize clients' self-esteem with

the definition in

mal way." This definition is perfectly adequate if we merely want to communicate the general topic

memory

when

necessarily meaningful or relevant. For example,

"to repeat or speak

we want

agree on

all

Notice that a highly operational definition

aloud from or as from memory, especially in a for-

Let us suppose, however, that

not

an

operational definition of students' ability to recite.

mensional.

Example

Notice that

if a

researcher

ability to recite is defined as the number of words mispronounced, missing, or misplaced when students repeat Christian Abzab's poem The Road Taken aloud from memory in front of a panel of three

theless

exist

because the researcher has to take

physical steps to conduct the research. That definitions exist but

teachers.

may be

Notice that the definition in Example

we might

is

1

not

because

erational definitions assist those

will the recitation take place (a class-

at

are trying to

is

if

helpful in

reducing spousal abuse, but the researcher

know

fails to

provide a highly operational definition, consumers

of research will not know at

who

research shows that a certain program

the students?"

important to note that no attempt

is,

not impos-

apply research in everyday settings. For instance,

or auditorium)?" "Will the teachers be male

It is

is difficult, if

Besides allowing for replication of research, op-

still

or female?" and "Will the teachers already

it

the

they

sible, to replicate the research.

have questions about the physical arrangements such as, "Will the students stand while reciting?" "In what fully operational

room

is,

When

unstated.

are unstated in a research report, the report best, vague. In addition,

room

provide opera-

tional definitions of variables, the definitions never-

The

type of

fails to

1

opera-

gram

how

to

conduct the pro-

in applied settings.

tionalizing will result in a completely operational definition because there are an infinite

number of

physical characteristics that might be addressed in

any definition (e.g., the humidity in the room, the level of lighting, the type of flooring, the color of the walls). Thus, instead of striving for completely operational definitions, we try to produce defini-

We have the most confidence in results of research that have been replicated by independent researchers because a given researcher may have blind spots, unconscious biases, etc. Also, a given researcher may have been 1

tions that are adequate to permit a replication in all

important respects by another researcher.

A

replica-

unlucky and have results that were influenced by large random errors. Independent replications by others reduce the odds that these factors are the cause of a particular

an attempt to confirm the results of a study by conducting it again in the same way. Of course, tion

is

there

is

much

subjectivity in applying this criterion

finding.

17

Exercise on Topic 1.

Which type of definition

2. In practice,

3.

indicates physical steps?

are operational definitions ever fully operationalized?

Which of the following

definitions of gregarious

A. Talking on the phone with friends for B. Being

4.

open and friendly when

at least

is more operational? two hours each week.

in social gatherings

with others.

Which of the following

definitions of being computer literate is more operational? two formal courses of instruction on the use of computers in an accredited school. B. Having knowledge of the origins and uses of computers in modern society and their implications.

A. Taking

5.

8

To

6. Is

at least

replicate the research

it

of others, do

we need

"operational" or "conceptual" definitions?

possible for an operational definition to be too narrow?

Questions for Discussion 7.

Suppose you read a research report claiming

that

low socioeconomic

status

(SES) children have lower

SES children. In the report, the only definition of self-concept is "feeling good How much credence would you give the results in light of the definition? What additional

self-concepts than high

about oneself."

information, if any,

8.

would you want about

the definition if you

were planning

number of times each work most mornings?"'

In a research report, job satisfaction is defined as "the

participant said 'yes' to

questions such as 'Do you look forward to going to

Is this definition

operational? If not, what

9. Is

is

narrow

in

terms of how

we normally

think about job satisfaction?

not?

10.

Write a highly operational definition of "success

11.

Write a highly operational definition of "motivation to succeed on the job."

For Students Name

in college."

Who Are Planning Research

the major variables

you

will

be studying. Define each, trying

(Note: After you have read published research on your topic, you fine

completely

missing from the definition?

the definition in Question 8 too

Why? Why

12.

to replicate the study?

some of your

to

be as operational as possible. to come back here and rede-

may want

variables in light of how other researchers have defined them.)

Topic 9 Quantitative vs. Qualitative Research:

The

of quantitative research are pre-

results

sented as quantities or numbers

(i.e., statistics).

phasize instruments that yield words, such as unstructured interviews or direct, unstructured obser-

In

qualitative research, the results are presented as

vations of police officers and their administrators.

discussions of trends and/or themes based on words,

not

statistics. In

addition to the difference in

I

When

how

deciding which

members of

the police

force to use as participants, quantitative researchers

number of charactwo types of research. To understand some of the major ones, consider this results are presented, there are a

would tend

teristics that distinguish the

possible within a limited research budget by objec-

research problem:

A

metropolitan police force

tive instruments

demoralized, with signs such as high rates of absen-

made

is

such as an anonymous, objective

questionnaire that takes

is

which

to select large samples,

little

time to administer.

would tend

qualitative researcher

'

A

to select a small

teeism, failure to follow procedures, and so on. Fur-

sample for the reverse reason.

thermore, the press has raised questions about the

research, a quantitative researcher

effectiveness of the force and

spend a small amount of time directly interacting with the participants (largely due to the nature of

its

sponse, the police commission

is

leadership. In re-

planning to em-

A

When

conducting the

would tend

ploy a researcher to identify possible causes and

the instruments used).

solutions.

the other hand, might spend a considerable

If a researcher with a quantitative orientation

qualitative researcher,

bers of the force over an extended period.

review of the

on

amount

of time interviewing and observing various

is

would probably begin with a research literature on demoralized

retained, he or she

to

mem-

the review, the researcher

While working with the participants, 2 qualitative researchers would be open to the possibility of mak-

would attempt

to

develop hypotheses to be tested by

ing adjustments in the instrumentation, such as re-

research. This

is

a deductive approach to planning

wording questions or adding questions based on earlier responses by participants. A quantitative researcher would seldom make such adjustments during the course of a research project. Instead, quantitative researchers would plan their research in detail in advance and follow the plan closely throughout the study because mid-stream deviations might be viewed as introducing subjectivity into the study. It

police departments.

the research. That

from the

From

is,

the researcher

is

literature possible explanations

deducing (i.e.,

hy-

potheses) to be tested. In contrast, a qualitative researcher

would tend

to use an inductive

planning the research. ple, in

He

approach

to

or she might, for exam-

begin to gather data on the specific police force

question by making preliminary observations and

conducting informal interviews. The resulting pre-

is

liminary findings might be used as a basis for plan-

searchers emphasize "objectivity," qualitative re-

ning what additional types of information to collect

searchers believe

and how to collect them. Thus, rather than approaching the research task with preconceived no-

inherently

on published theory and research, a would emphasize induction preliminary data that were collected.

qualitative researcher

that qualitative researchers (like quantita-

tive researchers) typically

lished literature

research reports.

all

observational processes are

subjective and open to

interpretation.

Because of this, qualitative researchers sometimes mention relevant details of their personal backgrounds (such as having a mother who was a police

tions based

from the Note

important to note that while quantitative re-

1

examine previously pub-

In addition, quantitative researchers usually attempt to

random sample

select a

and include reviews of it in their However, quantitative researchers

in

which

all

participants have an

equal chance of being selected. This can be done, for

example, by drawing names out of a

hat.

The uses of

use literature as the basis for planning research

random samples and

while qualitative researchers do not.

discussed in later topics.

When deciding what types of instruments (i.e., measuring tools) to use, quantitative researchers tend to emphasize those that produce data that can be easily reduced to numbers, such as structured

purposive sample of people she believes are key informants in terms of social dynamics, leader-

would tend

to

is

more

ship, etc. 2

Note

that the

term "participants" implies that the

indi-

viduals being studied have voluntarily agreed to partici-

pate in a given research project.

such as multiple-choice questions. In

contrast, a qualitative researcher

A qualitative researcher

likely to select a

questionnaires or interview schedules with objective formats,

their relationship to statistics are

When

individuals are

being observed without their consent, they are more

em-

likely to

19

be called "subjects."

officer) in order to inform readers of their research on possible sources of bias in collecting and inter-

ize the results to

preting the data.

qualitative researchers tend to limit their conclu-

Finally, quantitative researchers tend to general-

one or more populations, while

Also, quantitative researchers tend to summarize

sions to only the individuals

responses with statistics and seldom report on

studied.

all

Should

the responses of individual participants. Qualitative

researchers, als'

on the other hand, tend

commission

police

the

who were select

directly

a

re-

searcher with a "quantitative" or "qualitative" ori-

to cite individu-

responses (such as quoting individual partici-

entation? Criteria for

pants) in the Results section of a research report.

making such

a decision are

described in the next topic.

Exercise on Topic 9 1

Do

"qualitative" or "quantitative" researchers tend to rely

more on published research

literature in plan-

ning research?

2.

Which method of research

relies

3.

Which method of research

is

4. In

on the inductive approach?

more

likely to lead to a statistical report

which method of research would

a researcher

be more likely

to

of results?

make adjustments

to the interview

questions during the course of a research project?

which method of research and lengthy?

5.

In

6.

There are more likely of research?

7.

In

to

is

the interaction between the researcher and the participants

be quotations from participants

which type of research do researchers have

little

in the Results sections

more

intense

of reports on which type

interest in generalizing the results to populations?

Questions for Discussion 8.

In general, are

you more

likely to believe research results that are presented as

themes and trends ex-

pressed in words or results described with statistics? Explain. (If you have not read academic research extensively, consider secondary reports of research such as those found in newspapers, magazines, and

textbooks.)

9.

Do you your

believe both qualitative and quantitative research have valuable roles in advancing knowledge in

field

of study?

For Students 10.

At

Why? Why

not?

Who Are Planning Research

this point, are

you leaning toward conducting "qualitative" or "quantitative" research? Explain the you will be learning more about qualitative research in the next topic.)

basis for your choice. (Note that

20

Topic 10 Quantitative vs. Qualitative Research:

In the last topic,

we

finished with the question,

tative

Some of the

it

criteria that

making such a decision

When

D.

"Should the police commission select a researcher with a 'quantitative' or 'qualitative' orientation?"

potential participants are not available

for extensive interactions or observation, the quanti-

approach should be considered. For example, might be difficult to schedule extensive interviews with chief executives of major corporations.

should be considered when

are:

A. Some research questions inherently lend themselves more to the quantitative than the quali-

However, the chief executives might be willing respond to

a

brief objective-type

For example, "What

is

the impact

which would provide data

of terrorism on the U.S. economy?"

is

a question

statistics.

tative approach.

that

lends itself to quantitative research because

economic

II

When

E.

to

questionnaire,

be analyzed with

that can

time and funds are very limited, quan-

and indicators are usually described with numbers. On the other hand, "What is the emotional impact of terrorism on at-risk health care workers?" is a question that lends itself more to the qualitative approach than the first question because it focuses on emotional impact, which is more difficult to quantify. Note, however, that the second question could be examined with either

titative research

qualitative or quantitative research.

(such as legislators or funding agencies sometimes

factors

When

B.

little is

New

known about

should usually be

tive research

topics are constantly

such as

new

diseases like

emerging

SARS, new

search. In

find (i.e.,

is little, if

the snapshot approach.

to ask

little

or, at

incorporated into a qualitative research pro-

ject.

When someone

dience, one should,

in-

says,

"Just

the

numbers,

eration of

may

more

possible, start

by present-

qualitative aspects of the findings. in this criterion is the notion

Notice that implicit that

both qualitative and quantitative approaches

might be used in a given research project, with each approach contributing a different type of informa-

might find about a

when

ing statistics. This might open the door to consid-

topics,

tion.

lit-

we have been considering quanand qualitative research as though they are opposites. However, some researchers conduct re-

Up

exactly what

about a topic?) In con-

to this point,

titative

qualitative researchers could start with broad

trast,

be favored

least,

tions are unlikely to impress them. For such an au-

On new

know

numbers"

"hard

require

new educa-

difficult to write structured questions

can you

audiences

do), quantitative research should

employ the deductive approach deduct hypotheses from previously published

(How when you know

When

F.

any, previously published re-

tle-known topic.

narrow aspects of research prob-

criminal con-

research). Also, quantitative researchers it

suggested because quantita-

lems. Qualitative methods do not lend themselves to

difficult to

it

it is

please," themes and trends illustrated with quota-

absence, quantitative researchers

its

this is

can be used to provide quick, inexpen-

sive snapshots of

fields,

such as using the Internet for

structional purposes in classrooms.

there often

tive research

favored.

in all

cerns such as domestic terrorism, and tional techniques

types of research,

a topic, qualitainitially

might be favored. Although

an arguable criterion for selecting between the two

a blend of the two approaches. For

questions and refine them during the course of the

search that

interviews as various themes and issues start to

example, a quantitative researcher

emerge. Based on the qualitative results, theories might be developed from which hypotheses could be deduced and subsequently tested by using quan-

structured interviews to collect data, reduces the

participants to support the statistics,

titative research.

research that has

C. is

When

is

A

more

likely to

searcher

who

some of the

is

conducting

characteristics of both

As you can see, our hypothetical police commission needs to make a complex decision. How would

skilled qualitative researcher

you answer

willing to spend considerable time breaking

through the barriers that keep researchers out

uses semi-

approaches.

closed or secretive, qualitative research should

who

who

data to statistics, but also reports quotations from

the participants belong to a culture that

usually be favored.

is

this topic?

is

the question in the first paragraph of

What

is

the basis for your answer? For

more information on

be successful than a quantitative retends to spend much less time inter-

the characteristics of qualita-

tive research, consult Part

acting with participants.

21

H

of this book.

Exercise on Topic 1.

1

lends itself more to quantitative research?

Which of the following

How

do the social relations of adolescents who use illicit drugs differ from those who do not use them? B. How do school attendance and grades earned in school differ between adolescents who use illicit drugs and those who do not use them? A.

2.

Which of the following lends itself more to qualitative research? A. What are the differences between the social interactions of college students on commuter campuses and students on campuses where most students live on campus? B. To what extent does family income predict whether a student will choose to attend a commuter college or a college where most students live on campus?

3.

Suppose you want to do research on members of a secretive fraternity. According to the information this topic, which type of researcher is more likely to gain access to the participants?

4. If little is

5.

known about

a

new

topic,

which type of research

is

recommended

in

for initial use?

For which type of research must participants usually be available for extensive interactions with

re-

searchers?

6.

Which type of research

is

more

suitable for getting a quick snapshot of a

problem?

Questions for Discussion 7.

How would

8.

Suppose a team of researchers wants

whom

the

two questions to

in the last

paragraph of this topic?

conduct research to identify the characteristics of professors

students perceive as being excellent.

search?

9.

you answer

Would you

advise them to do qualitative or quantitative re-

Why?

Name a problem

in

your

field

of study

that

would probably lend

itself

more

to the quantitative than the

qualitative approach.

For Students 10. In light

Who Are Planning Research

of the information

in this topic,

have you changed your mind about your answer

the Exercise on Topic 9? Explain.

22

to

Question 10

in

Topic

Program Evaluation

1 1

Consider a school that receives a foundation new program that emphasizes parental involvement and shared decision-making. In this

that the teachers'

grant for a

program, decisions are

made by an

sired,

more of

administrator

The ultimate purpose

improve

in a variety

is

to help students

the

and effectiveness of the programs

for

to give parents it is

course

in the nature

of an

to

of a treatment during

experiment,

skilled

program

evaluators are open to such modifications. In fact,

program evaluators collect information during the course of a program that assists in the process of modifying the program while it is being imple-

almost always require a report on the implementation

not as high as de-

a voice in decision-making. Although

make changes

of academic and social skills. most others, granting agencies

In this case, as in

is

may be made

almost unheard of for an experimental researcher

with the advice and consent of both teachers and parents.

involvement

adjustments

which

To prepare such a report, reprogram evaluation by engag-

they provide funding.

mented. Collecting this information

searchers conduct a

researchers are attempting to understand underlying

mative evaluation. Formative evaluation has two prongs. First, information is collected on the process of implementing a program. For example, when looking at the process, we might ask, "Were the parents notified of the program in a timely manner?" and "Were the proposed meetings of parents and teachers conducted?" These questions clearly ask about the process, not the ultimate goals of student improvement. The second prong of formative evaluation involves collecting information on the progress toward the ultimate goals. For example, periodic achievement tests might be administered to see if students are showing signs of improvement. If not, evaluators and program administrators might rethink the process they are implementing and make

theories that explain behavior without necessarily

appropriate changes.

looking for direct applications.

responsible for the program can often prevent disappointment in the final results of a program

ing in evaluation research.

At

first

glance,

might appear

it

we

that

should

conduct experimental research, which, as you recall

from

give treatment(s) (in this case, to

may

which we a program) in order

earlier topics, is a study in

observe their effects. Indeed, while elements of

evaluation research resemble experimental work, there are First,

some major

differences.

program evaluation

plied research

(i.e.,

is

almost always ap-

research in which

we wish

to

apply the findings directly to such practical decisions as whether to continue funding the program

and whether to modify it). Experimental research, on the other hand, is often basic research in which

Second, new programs are, or should be, based on a needs assessment. A needs assessment is nonexperimental research in which researchers attempt to determine the practical needs of those who will be served by the program. For a school-based program, we might ask questions such as, "What types of skills do the students need to acquire?" and "What types of program interventions to promote these skills will be most readily accepted by students, parents, and teachers?" Pure experimental research is seldom preceded by a formal needs assessment, which is associated with program evaluation because a major focus of evaluation is to estimate the extent to which a program has met the

By

looking

at

is

called for-

progress, those

evaluation.

When

evaluators collect information about par-

ticipants' attainment

of the ultimate goals at the end at the end of a school year),

of the program (such as the

activity

called

is

summative evaluation.

A

summative evaluation report contains information about the final or long-term benefits of the program for

its

ultimate clients (such as students).

Summa-

comparison with a control group. For example, students in a program might be compared with similar students in other schools who are not in the program. Note that while

tive evaluation often involves a

experimental researchers typically strive to use random assignment (like pulling names out of a hat) to

needs revealed in an earlier needs assessment. Third, the programs, which are analogous to

form experimental and control groups, program evaluators usually have to find an external group

treatments in an experiment, are usually subject to

(such as students at another school) to serve as the

change during the course of the evaluation. For example, perhaps a program is designed to give teachers a major role in decision-making, with only a minor role for parents. If, in midstream, it is found

control group.

As you can

23

program evaluation is a comform of applied research.

see,

plex, specialized

Exercise on Topic 1

Is

"program evaluation" or "experimental research" almost always applied research!

2. Is

a needs assessment associated with "experimental research" or

3. Is

it

4.

1

"program evaluation"?

acceptable to modify the treatments (programs) during the course of a program evaluation?

Suppose, as part of a program evaluation, an evaluator asks, level

by the end of the program?"

Is this

"How many

children were reading at grade

question relevant to "formative" or to "summative" evaluation?

5.

Suppose that as part of a program evaluation, an evaluator asks, "Are the clients in the job-placement program writing better resumes?" Is this question relevant to "formative" or "summative" evaluation?

6.

Suppose, as part of a program evaluation, an evaluator asks, "Were key program personnel hired on time?"

7.

Is this

When we

question relevant to "formative" or "summative" evaluation?

look

at the

process of implementing a program, are

we

conducting "formative" or "summative"

evaluation?

8. Is

examining program participants' progress toward attaining the ultimate goals of the program part of

"formative" or "summative" evaluation?

9. Is

the attainment

of the

final goals

of a program by participants a topic for "formative" or "summative"

evaluation?

Questions for Discussion 10.

1 1

Suppose you were on the board of a foundation that was giving a grant for a program to a social welfare agency. Would you prefer to have the program evaluated by an employee of the program (such as the program director) or by an external, independent evaluator? Why? Sometimes, government agencies and organizations fund certain programs despite prior negative summative evaluations. Speculate on some of the reasons for this. Are any of the reasons justifiable?

For Students 12.

Who Are

Planning Research

Will you be evaluating a program in your research? If yes, will

name

be conducting both "formative" and "summative" research.

24

the

program and indicate whether you

Topic 12 Ethical Considerations in Research When

planning research,

harm

sider potential

from

it is

imperative to con-

to participants that

might

purpose of a study sometimes makes it difficult or impossible to conduct research on important topics. For instance, suppose researchers want to study the

result

their participation. Often, certain treatments

that researchers

might want

to administer in ex-

influence of lobbyists on a group of state legisla-

perimental studies present hazards. For example, a research psychologist might expose an experimental

tors.

group to an anxiety-provoking stimulus in a study designed to advance a theory of the sources and

them around unobtrusively)

The researchers might get some

legislators to

allow the researchers to "shadow" them if the

(i.e.,

purpose

vaguely described as "to understand the

follow is

only

state's

po-

of anxiety. Clearly, some participants might suffer mental anguish as a result of being exposed

(especially those

to the treatment.

duly influenced by lobbyists) would agree to being

effects

Participants

litical

might be harmed in nonexFor example, the process of ex-

also

perimental studies.

shadowed

might cause participants to focus on them again, leading to renewed anxiety, sleeplesssive parents)

and so on. Because of such potential problems, the research community has developed a body of ethical values regarding the use of humans as participants. The primary value is that participants must be protected ness,

it

is

not always possible to anticipate

all

when using new

treatments or measuring tools. Because of this, most

and large school districts have research committees that review research plans for potential universities

harm

legislators

allow themselves to be un-

if researchers reveal the true

purpose?

Is

ethical to present only a

1

from both physical and psychological harm. Unforthe potential for harm, especially

who

how many

vague general purpose that does not reveal specific goals? Do government employees have a right to privacy on the job? These types of questions illustrate the difficulties in balancing the need to protect participants with the need to collect information of benefit to society. A key to promoting these values is informed consent. To use informed consent, we inform the participants of (1) the general purpose of the research, (2) what will be done to them during the research, (3) what the potential benefit(s) to them and others might be, (4) what the potential for harm to them might be, and (5) the fact that they may withdraw at any time, even at midstream during the it

ploring sensitive traits (e.g., relationships with abu-

tunately,

process." However,

research, without penalty. This information should

be provided

to participants.

Another important value is that participants have a right to privacy. For example, most would probably agree that it would be a violation of parents'

in writing,

and the participants

(or their

guardians) should sign the informed consent form to indicate that they understand

Another key

to

it.

promoting the

ethical values

we

rights to privacy for researchers to question children

are considering

about discord between their parents without paren-

participation in a study. Debriefing consists of re-

consent, even if the results might be very useful

viewing the purpose(s) of the study and the proce-

tal

and others. that participants have a right

dure^) used as well as offering

to educators, sociologists, psychologists,

A to

related value

is

debrief participants after their

is to

have the data collected about them as individuals

able.

when

to share the results

become availThe process of debriefing should also include

with the participants

the results

kept confidential. Even if participants freely and

reassurances that the data will remain confidential.

knowingly provide information to researchers, the researchers have an obligation not to disclose the information to others unless the identities of the participants are disguised or hidden, which is often accomplished by using statistics such as group

In addition, participants should be allowed to ask for information about

which they

searchers should try to identify participants

may need more harm

averages.

Also, almost

any aspect of the study in During debriefing, re-

participated.

to

who

help in overcoming unanticipated

them than a standard debriefing session

provides.

all researchers agree that partici-

pants have a right to knowledge of the purpose of the research before they participate. Having this

1

knowledge, they are in a better position to determine whether they want to participate. Unfortunately, complete honesty with participants about the

For more information on

this topic, refer to

Ethical

Principles in the Conduct of Research with Human Subjects, published by the American Psychological Association. Visit

25

www.apa.org

for

more information.

Exercise on Topic 12 1

Should researchers take steps to prevent psychological harm as well as physical harm

2.

Should participants be told that they are free to withdraw from a study

3.

Under

the principle of informed consent,

is it

at

to participants?

any time without penalty?

acceptable to hide the general purpose of a study from the

participants?

4.

Should informed consent be

5. Is

in

writing?

debriefing done "before" or "after" a study

is

conducted?

6.

What does

7.

Should information about participants be kept confidential even

debriefing cover?

if the participants freely

provided

it

to

researchers?

Questions for Discussion 8.

How would you • Is •

9.

10.

it

Do

answer these two rhetorical questions that are posed in the topic? vague general purpose that does not reveal specific goals? government employees have a right to privacy on the job? ethical to present only a

Suppose a researcher wants to keep a class of third-grade students in from recess to administer an attitude-toward-school scale. The purpose is to help teachers understand their students' attitudes and how they might affect students' achievements. Is there potential for harm in this case? Would it be wise to seek informed consent from the parents? Why? Why not?

A researcher interviewed adolescents on their possible use of marijuana (with their informed consent). During the course of the interviews, some participants named other individuals who use marijuana but who have not provided informed consent to the researcher. Does this raise ethical concerns? What, if anything, can the researcher do to protect the other individuals?

1 1

Suppose one of your instructors asked you to be a participant in a research project but did not tell you the purpose of the research. Would you ask for information on these points before deciding whether to participate? Would you feel pressured to participate because the researcher was your instructor?

For Students 12.

Do you

Who Are

anticipate that the study

measures will you take 13.

Planning Research you are planning has the potential harm?

to

harm

the participants? If yes,

what

to mitigate the potential for

Will you be obtaining informed consent? Will you have your consent form reviewed by your professor?

By

a university committee?

By

others? Explain.

26

Topic

A

theory

is

1

The Role of Theory in Research

3

a unified explanation for discrete

question, causing theorists to consider reformulat-

observations that might otherwise be viewed as unrelated or contradictory.

One of

studied theories in learning It

is

the

ing

most widely

it

account for the discrepancy.

to

Another major function of research is to provide the observations and conclusions on which we can induce theory. That is, we try to develop theories

reinforcement theory.

defines positive reinforcement as anything that

increases the frequency of a response from an ani-

that explain events

mal or individual. For example, most of us have given praise as a reward to a dog for sitting. To the

practice qualitative research often refer to this as

extent that the praise increases the sitting behavior,

observations).

it

as evolutionary. That

At first, reinforcement theory sounds obvious, and in a way, it is self-defining. So why has it been so carefully studied? Because it explains many apparently contradictory observations. For example, suppose at first we praise a dog regularly for sitting and after a while become lax and offer praise intermittently. Common sense might tell us to expect the sitting

behavior to decrease with the decrease

in

praise.

However, we might actually observe an

in-

ing the process of regularly revised

intermittent

many

circumstances,

reinforcement

more

1

is,

fied set

is

why

of principles that help explain

we would

you

aspect of

it

makes a

understanding of

a uni-

functions of research

is

to test

When

will find

it

easier to de-

show

thinking about theory as a basis for re-

is

mind

universal

no theory of human be-

that

(i.e.,

there almost always are

exceptions to the rule). This

is

why we

usually ex-

should ask for less feedback about test items than those who have

amine trends across groups in order to test or develop theories. However, do not overlook the possibility of designing a study specifically to examine those individuals who do not perform as predicted by a theory. Understanding how the dynamics of their behavior differ from those who act in the way

study that confirms this hypothesis

predicted by theory

level

of cognitive

we might deduce

that

engagement.

when

students

they will be tested again on the same mate-

rial,

those

who have lower

ting

70%

right)

goals

(e.g., a

goal of get-

wrongly answered

lends support to the underlying theory. is

you

your study has implications for validating and refining an important theory.

havior

the study

aspects of behavior related to

all

introduction to your research report if you can

axiom of self-regulated

A

some

that

learning theory states that the goals students adopt

2

is

fend your selection of a research topic and write the

it,

the theory. For example, an

higher goals.

it

observations

potential contribution to our

the theory. In addition,

certain

that are consistent with

this,

new

are looking for a research topic for a the-

we deduce hypotheses

From know

evolves) as

or term project, you would be well advised to

search, keep in

their

(i.e.,

consider a theory of interest to you. Testing

observe behaviors that do not seem to be

determine

developed dur-

desire by scientists to develop theories on major aspects of human experience should not

If sis

hypotheses derived from an existing theory. To do this,

is

lated "facts" collected via nontheoretical research.

consistent or related to each other.

One of the major

usually

be surprising. Unified explanations of phenomena are clearly more useful than a collection of unre-

under

behaviors increase in their frequency. Without

is, it

making observations, and

The all

effective than consistent

reinforcement. Thus, reinforcement theory

who

warrant.

crease in sitting because reinforcement theory indithat

observed. Those

grounded theory (i.e., theory that is grounded on 3 Grounded theory is often thought of

constitutes positive reinforcement.

cates

we have

may

help in refining a theory to

take account of exceptions.

Assuming

methodologically strong, failure to con-

firm a hypothesis calls the theory (or parts of it) into

1

In

this

discussion,

the

technical

3

For more information on the use of the grounded theory approach to research, see Topic 65 in Part H of this book. For an example of a study that employs grounded theory, see Noonan et al. (2004). Challenge and success: A qualitative study of the career development of highly

terms for various

schedules of reinforcement are not discussed. 2

For more information on this theory and to see an example of research based on the theory, see Azevedo, R., & Cromley, J. G. (2004). Does training on self-regulated learning facilitate students' learning with hypermedia? Journal of Educational Psychology, 96, 523-535.

achieving women with physical and sensory disabilities. Journal of Counseling Psychology, 51, 68-80.

27

Exercise on Topic 1

How

is

2.

Do we

3.

What

4. If a

1

a theory defined in this topic?

use "induction" or "deduction" to derive a hypothesis from a theory?

are the

two major functions of research mentioned

hypothesis derived from a theory

is

in this topic?

not confirmed, what implications does this have for the theory?

5. Is

grounded theory based on "induction" or "deduction"?

6. Is

the use of grounded theory

more

likely to

be associated with "qualitative" or "quantitative" research?

Question for Discussion 7.

Examine

the discussion of a theory in a textbook in your field.

search that supports

it?

Does he or she suggest unresolved

Does

the author of the textbook cite re-

issues relating to the theory that might be ex-

plored in future research? Explain.

For Students 8. Is

Who Are Planning Research

the purpose of your research to test a hypothesis

deduced from a theory? (This

is

tive

done by between qualita-

typically

quantitative researchers. See Topics 9 and 10 for a brief explanation of the differences

and quantitative research.) Explain.

the purpose of your research to make observations on which a theory may be built? (This is typically done by qualitative researchers. See Topics 9 and 10 for a brief explanation of the differences between qualitative and quantitative research.) Explain.

9. Is

10.

Will you be conducting research without reference to theory? (Note that

it is

possible to plan and conduct

research on practical matters without explicit reference to theory. For instance, suppose your local

schools went on double shifts because of overcrowding, and you want to investigate the effects of the

on students' attitudes and achievements. While there may be several theories that relate to you might conduct your research as a practical, nontheoretical matter in order to better inform decision-makers and taxpayers on this matter.) If you will not be referencing theory, will the results of your research still have important practical implications? Explain.

double

shifts

this situation,

28

PartB Reviewing Literature Reviewing the research conducted by others search project. In this part, it

will help

you

we

is

often the

first

step in planning a

will first explore the reasons for reviewing literature

identify a suitable idea for research as well as help

search tools and methods. Second,

we

will

examine how

you

new and

re-

how

select appropriate re-

to locate literature electronically.

that you already know how to locate books and articles in massand newspapers, the emphasis here is on how computerized databases can make your search of academic journals more efficient and precise. Finally, we will con-

Because

it

is

assumed

circulation periodicals

sider

some

basic principles for writing literature reviews, including

29

how

to cite references.

Notes j

1

(2c«.

«^>j

vc \

30

^

•

-^

Topic 14 Reasons for Reviewing Literature

If

you are planning your

first

research project for

a class assignment or a thesis, start

such an insight

^

by identifying

at the

beginning of his or her profes-

sional career.

broad problem area. This could be a practical prob-

In addition to helping

you

identify testable hy-

lem such as "the education of bilingual children" or "dependence on welfare," or it could be a more theoretical one such as "How valid is attribution theory in informal learning situations?" The second step is to review literature on the topic. Examining both the theoretical and research literature on a topic will usually help you identify a testable hypothesis of limited scope (limited so that it can be tested within a reasonable amount of time and with the resources you have available). If you are having difficulty identifying a research hypothesis, you might consider replicating a 2 study that has already been published. In a strict replication, researchers try to mimic the original study in all important respects because the purpose is to see if the same types of results will emerge. Of

potheses,

course, a strict replication should be undertaken

likely to struggle with both the organization

other benefits. First, you

Note

that the findings

of a study

academic community as

were used successby other researchers and also avoid those found to be seriously flawed. Secondjyou may be

able to

overstated.

the

be required

ignore this advice are

and

dissertations,

to present

your

literature

review

will in a

introduction.

show how your

by established researchers. Finally, your review of research allows you to demonstrate to your instructors, thesis committee members, and others that you were able to locate research relevant to your hypothesis, evaluate the literature, use it in planning your research, and cite

ex-

it

appropriately in your review of literature. In the next topic,

have

will consider

vide

we

will explore

ing,

acceptable in

of the objectives for the course.

31

how

and

to organize

it.

in

how

to search

Topic

16,

we

Topic 17 will pro-

some important suggestions for effective writand Topic 18 will show you the basics of citing

references.

can be tested through direct observation. is

is

research flows from important research conducted

See Topic 7 to review hypotheses as well as purposes and questions, which are alternatives to hypotheses, as the heart of research. A testable hypothesis is one that to see if this

you probably

the importance of your topic and to

not a direct extension of exist-

Check with your professor

it

crafted review of research

1

light

writing a

Note that a wellshows your readers the context within which you were working. It can also help to justify your study if you use it to_establish /

that contains the

for literature electronically,

2

When

separate chapter immediately following the chapter

ing research. While there are important cases of this

will

who

to cite relevant research.

tion. In theses

arrive at a

is

Students

typical to integrate the literature into your introduc-

them.

It

large

research report for publication in a journal,

Published reviews of research often point

of science, they are rare. tremely unlikely that a novice researcher

research

and wording of their reports and, despite their best efforts, produce a sophomoric report. When you write a research report, you will be

out such conflicts and offer possible explanations of

the history

how

The importance of reading

numbers of original research reports before attempting to write one's first research report cannot be

of published research are conflicting, ajhjrdj possibility is to plan a study designed to resolve a

in

fruitful). Third,"

read just for content but also notice

sults

is

for

to write research reports

reports are written.

measurement technique). Notice, though, that if the study fails to replicate, you will not know whether it is because the original study was in error or because your modifications changed the nature of the study. Because there are many topics on which the re-

creative idea that

how

expected

may

one of your ideas

(i.e.,

already been thoroughly investi-

you can by paying careful attention to the style and organization used by authors of published research. To do this, do not learn

invalid.

After reviewing research, you

may have

gated and shown to be not

Another possibility is to locate an important study and conduct a modified replication (i.e., a replication with some major modification(s) such as examining a new population or using an improved

conflict.

avoidMead ends

research

that others

by

identify measuring

fully

only on studies with potentially important implicaare not able to replicate are usually dismissed

may

tools (called instruments) that

1

tions.

reviewing published research provides

Exercise on Topic 14 1

2.

When

planning research, should you

first

identify a broad

problem area or develop a testable hypothesis?

Suppose you conducted a replication of a study but used a sample from a population different from the one used in the first study. What type of replication did you conduct? ^ oJ-« £ tcjfro^ ?

3.

Should you be surprised

4.

"According

if

you

find conflicting results in the research literature

to this topic, students

would be wise

to try to find a creative,

new

on a given topic? \J