Stimulus differentiation by schizophrenic and normal subjects under conditions of reward and punishment

381 75 4MB

English Pages 77

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Stimulus differentiation by schizophrenic and normal subjects under conditions of reward and punishment

Citation preview

m m m m s m w m m w m by as© w m m m

A dissertation submitted 1% partial fulfillment of the raqairoBeats far the degree of Sector of Phlle** •epfcy, la the Department of Psycfeology, Is the Graduate College of the State Uaireraity of Iowa Ithrsary, 19JS

S tate Univ -jt sow® LibHAr-. hi^

ProQuest Number: 10598589

All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is d ep en d en t upon the quality of the copy submitted. In the unlikely event that the author did not send a com plete manuscript and there are missing pages, these will be noted. Also, if material had to b e rem oved, a note will indicate the deletion.

uest, ProQuest 10598589 Published by ProQuest LLC (2017). Copyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States C o d e Microform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106 - 1346

T \e\bo Co^)i 9s.

I would liltw to express ay sincere gratitude to fir. & H. Sedniefc and fir. X. S. Farter for their always helpful and stimulating criticism and advice throughout the course of the investigation. Mr. Austin W. Berkeley of the Worcester State Hospital has helped lfcraeasur&bly with the construction of the apparatus and the solution of numerous technical problems. . My wife, fidith, has been a constant source of encouragement. X m deeply grateful to her.

ill

TABLS

OF

COHBiiTS

Chapter !• II. III.

fag# Introduction . . . . . . . Fnblcft • t . « • • «

..........

......

..............

?

£xpe?lment&l F z » a « d m ...............

10

Sdbjeota

If.

1

.

10

Apparatus......................

11

Pxecedure..............................

13

Braatft.....................................

23

S a m i Only - Saesion 1 .................................

Stk

Bward^Puaishatent.............

SI

Bsward Only ▼#. Beward-Punishment •

...............

Barnard Only « Session 2

35 •

39

Cenparlaea of F irs t and Second Conditions U3

Howard O n ly .............................. f. fl.

Biacusslen.............................

H5

Summary.......................

56

S i f t m o t i ...................... Appendix A .................. Appendix B Appendix

6 ..................

So .. ......................

. . . . . . . . . .

6g 6H 6S

!▼

t a b u

o w

n s r s s s

1

Sublet1* Vie* of Apparatus

* . . * * * ........ ♦

£

i £ ^ l Ifipnt dl Apparatus

. . »

3

Mean frequency of Poll £»*pen»ea to Jfedi Stimulus By Sehisspfcreaio m € Normal SubJoeto Under Initial IfRavt OxaXy Condition

2?

Mean frequency of fuXX Bespoasee to Sadb Stimulus ly Schisep&renic and normal Subjects Under Seward* lition

3D

la £

?

........

. •

of Pull Besponoeo to Sack Stimulus Isepbrenic and ^ hbbI Subject# Under Condition * * • ......... . . * .

If

3^

of Puli Beepenses to Sack Stimulus By Schisophrenic and Normal Subjects Under Bewsrd Only andBeward-Puui^meat Conditions ***«.**.

3?

Uses frequency of Ball Hesponce© to Each Stimulus and Somal Control Subject* Under Only Condition » * * * . • * • • » « *

1*1

ef full Responses to Eadi Stimulus t

S A B It S

0 F

I A B U S

fefels 1 ft

Iff

IF

T

VI

Fags ........

2g

Mass* and S.D. •» o# Distributions of Fall Responses to Sacla Stimulus jy Four Group* Under First Haward Only Condition « . .........

25

Moano and S.B. *s of Distribution# of Fall Eesponse# to Sach Stimulus £y SX-F and 22S*»? ........... •

. .

29

Means and SaD.'* of distributions of Ball Besposse# to Stimulus by Matebod SX-F ........ . . * . • said 2PHP Groups

33

Moans and $•$•** of distributions of Fall Besponse# to Saeh Stimulus for S0-S2 end U0~B2 Groups » » . » » » . ' « . » ............

No

Strength of legati** Full faadsaoy to FaeM Stimulus by Schtaopbsmle and Bormal Subjects Under tbs Bsssaid^Funishmant Condition..........

%

Summary of tbs Krperimental Design • . .

1

numerous studies have attested to the loss of efficiency manifested by schizophrenics la respect to various intellectual and meter processes. fhis lose or *psych*legical deficit** Is evidenced this term has been suggested by Hunt and Gofer, end is defined as a condition which Is manifested *whea say person perform* la some situation at a level of efficiency below that expected from compari­ son with typical individuals or from seme indicator in his own pro* seat or past behavior** (Id, p» 971)* Vhe use of this term in characterising the pathognomonic aspects of schizophrenic behavior appears preferable to such previously used concepts as *&«®©nti&,* *&eterioration* and Bregression, * because it Is net associated with either a particular nosological and descrip­ tive psychiatry or a specific theory of personality development* by schisophrenic patients in such diverse areas as psychometric per­ formance (1*1, 19)* speed and accuracy of tapping (£2), learning of motor tasks (12, 15) > reaction time and choice reactions involving simple visual discrimination (11). Various psychiatric writings suggest that another (and pezS&aps basic) area of deficit in schizophrenic patients involves a relative inability to make differential responses to different stimuli within a given class*

Per instance, the Withdrawal* be­

havior so often represented as characteristic of the schizophrenic syndrome (1* 10) may be interpreted as a general, non-selective avoidance tendency, involving little or no differentiation among specific persons or events*

Thus Schildev stalest

2! A further type of schizophrenic reaction is the levelled and diffuse defense against everything which is done to him* ..♦It is characteristic of such reaction that generally every­ body who approaches the patient is reacted to in the same way® The thought and language process of the schizophrenic is ar­ rested and the individual tries to reach a satisfaction In a world of incomplete differentiation {21 , p® 38£)® Precise, quantitative evidence in support of this type of psychiatric observation of the failure of schizophrenics to differen­ tiate among stimuli along a given dimension is meager and largely indirect.

Bender and Schiller (2) attempted to establish a condi­

tioned finger withdrawal to a tactile stimulus in 16 stuporous cata­ tonic schizophrenics, using a strong shock as the unconditioned sti­ mulus® After the completion of conditioning, they noted that the conditioned response could be elicited not only by the original conditioned stimulus (a touch on the forehead) but also by touching distant points of the skin*

They concluded that schizophrenics

tend to show an Mincomplete analysis of the situation** (2 , p» 362)® Unfortunately, the failure of these experimenters to secure data on normal control subjects rendered their results inconclusive with re*** spect to the relative ability of these patients to differentiate among stimuli as compared with normal subjects® In a preliminary investigation related to the present study, the writer compared the ability of schizophrenic and normal subjects in differentiating among visual stimuli.

The stimuli con­

sisted of 12 standard geometric forms originally used by Gibson (6 ) plus four variations of each standard form®

Bach of the four varia­

tions was ranked by 2$ normal subjects in terns of degree of simi-

larity to its standard*

The variation of each standard judged most

similar was designated a Class X design, th® next most similar, Class II, etc. The 12 standard foxms were presented in a predetermined order to Ik normal and 11 schizophrenic subjects, each design being exposed for on® second with a four-second delay between exposures* After four trials the variations as well as the standard forms were presented in random order, and th® subjects were asked to Indicate whether or not they recognized the form as one which had been shown to them previously*

There was no significant difference between the

^recognition* scores of the schizophrenic and normal groups on th® 12 standard forms| the schizophrenics, however, had higher *recogaition* scores on all four classes of variations, the differences be­ tween the groups at each level of similarity being significant at the $ per cent level of confidence*

It could therefore be concluded

that schizophrenics are reliably poorer than normal subjects in differentiating among stimuli under the conditions of this experi­ ment* There is an increasing amount of experimental evidence in­ dicating that motivational factors may be significantly involved in the production of this and other kinds of performance deficit in schizophrenics*

Wittman @ 6) for instance has reported high correla­

tions, averaging better than +*60, between ratings of cooperative­ ness in schizophrenic subjects and their performance scores on reason ing and memory tests*

Huston and Shakow (12,13) found that patients who

4 showed greater cooperation on various motor tasks tended to reach performance levels more closely approximating those of normal sub­ jects*

They concluded that a deviant motivational state was on© of

the factors leading to the generally poorer performance of schizo­ phrenics® As MoGeoch 07) has stated, it is usually assumed that a subject* s cooperativeness in meeting the demands of any test situa­ tion is, in part, a function of certain secondary (social) motiva­ tional systems (efJS)*

This suggests that the malfunctioning of

such acquired drive systems may underlie the poor cooperation so often manifested by schizophrenic subjects in experimental tasks* Recently, Cohen (U) has demonstrated that differences in the performance of normal and schizophrenic subjects are more pro­ nounced when the important motivating condl tions are of a secondary or social nature than when they are primary*

This investigator had

half the subjects in both normal and patient groups learn several serial motor tasks under conditions of primary motivation induced by electric shock, while the other half of each group learned under conditions of ordinary rapport, in which secondary motives would be of major importance*

The normal subjects performed in the same man­

ner under both test conditionsj schisophrenic subjects, however, behaved differently under the two conditions*

Under primary motive?*

tion the patients* performance like that of the normal subjects showed continued improvement in learning throughout the session, whereas the "rapport* group of patients showed an eventual performance

5 decrement following a temporary improvement in the task*

These

findings indicated that schizophrenic patients behaved less like normal subjects under conditions in which performance depended more upon the effectiveness of secondary drive than under condition® in which a primary drive was operating# This supposition concerning the inadequate functioning of secondary motivational systems in schizophrenics is supported by the observation (9) that such patients are likely to engage in simple laboratory tasks anticipating and accepting defeat, preoccupied with extraneous stimuli or entirely indifferent to the outcome of their performance. It does Hot seam unreasonable to suppose that this kind of deviation from the motivational pattern generally typical of nor­ mal individuals may underlie the generalized withdrawal behavior which has been described as characteristic of schizophrenic patients# It is important to note, furthermore, that psychiatrists have fre­ quently considered such withdrawal behavior to be a function of the degree of stress or threat in th® situations to which the patient is exposed# Schilder, in discussing the failure of schizophrenics ade­ quately to differentiate among cues in th© environment, states $ It is as if the schizophrenic would say, ”This fully differentiated world of yours is much too difficult and dangerous and therefore I content myself with a more primi­ tive world” ( 21, p. 389)*

Eodnick^ has demonstrated experimentally the differential

A

* Unpublished investigation conducted at the Worcester State Hos­ pital, Worcester, Massachusetts. effects on normal and schizophrenic subjects produced by the inter­ polation of stressful stimuli during performance on an airplane type pursuit meter. Whereas normal subjects learned to respond different tially to cues associated with stress and non-stress trials and ex­ hibited continued improvement in th© task on the latter trials, schizophrenics were unable to differentiate between the two condi­ tions once they had experienced stress and showed no subsequent im­ provement in learning after stress stimuli had been interpolated# These findings indicate that motivational processes in schizophrenics are particularly likely to deviate from the normal under conditions of environmental stress# It has been suggested that th© withdrawal behavior which characterizes schizophrenics in­ volves a lessened ability to differentiate among cues in the environ­ ment, and that such withdrawal tendencies are exaggerated under con­ ditions of stress*

If these suppositions are correct the ability of

schizophrenics to differentiate among environmental cues might be expected to be more adversely affected under conditions involving social punishment than under less threatening condL tions#

7 Chapter n problem The present study was designed to determine whether the assumed lessened ability of schizophrenic subjects to differentiate among stimuli within given classes would be evidenced in a simple experimental situation involving auditory stimuli5 and whether such inability would be accentuated under conditions in which secondary motives of avoidance of social disapproval and threat of censure were operating* These two suppositions were tested by comparing the per­ formance of schisophrenic and normal subjects under two test con­ ditionss 1# Reward Only condition Following practice on a training stimulus^, subjects were

In the interest of terminological consistency which subjects were instructed to respond during ferred to as the "training stimulus*" Any other pitch dimension is referred to as a "generalised

the stimulus to training is re­ stimulus along the stimulus.11

tested with generalized stimuli as well as the training stimulus* Correct responses to the training stimulus were rewarded by the lighting of a box indicating "RIGHT*" There were no after-effects for similar responses to th® generalised stimuli* 2* Reward-Puniahment condition Following practice on a training stimulus, subjects were

8 tested with generalized stimuli as well as the training ^^dmulus* •v*-:v •

As in the Reward Only condition, correct responses tq-'Jthe training stimulus were rewarded by the lighting of a box indicating "RIGHT.H Similar responses to the generalised stimulus farthest removed from the training stimulus was punished by the lighting of a different box indicating "WRONG." It was assumed that the patient group would show poorer differentiation, as evidenced by a flatter response gradient to stimuli under the Reward Only condition than would a comparable group of normal subjects* In the case of th® normal subjects, it was assumed that the Rewardr-Punishment condition would lead to increased ability to differentiate among the training and generalised stimuli and yield a steeper response gradient than the Reward Only condition, since punishment should lead to an avoidance of the prohibited response to the extreme generalized stimulus and also prestsnably to those stimuli more similar to the extreme generalised stimulus® If, however, the threat of punishment in the Reward Punishment condition has the effect, as has been supposed, of pro­ ducing withdrawal responses in schizophrenics, their inability to differentiate among stimuli should be exaggerated under this condition* On this assumption, the differentiating ability of schizophrenics would not be improved under the Reward-Punishaent condition.

Con­

sequently, the differences between schizophrenic and normal subjects should be greater under this condition than under that of Reward Only.

9 In summary* it was predicted that* 1» Schisophrenic® would show a flatter response gradient to genera­ lised stimuli under the Reward Only condition than would a comparable group of normal subjects3 2* The introduction of punishment for responses to the furthest generalised stimulus would steepen the response gradient for normal subjects, but would be less effective in steepening the gradient of schizophrenic subjects; 3* The greatest differences between the response gradients for schizophrenic and normal subjects should therefore b© produced under the Rewardhpunishaent condition*

10 Chapter III Experimental Procedure Subjects Four groups of subjects were used* an experimental group of 21* schisophrenic patients (SX)j a control group of 16 schizophrenic patients (SC); an experimental group of 21* normal subjects (NX)5 and a control group of 16 normal subjects (NC)• Schizophrenic Groups These subjects were male patients selected from the popula­ tion of two mental hospitals in which diagnostic practices are essen­ tially alike.

Twenty-six were patients at the Worcester State Hos­

pital* ?farcester, Massachusetts and 11* were from the Grafton State Hospital, Grafton* Massachusetts. All patients selected had been diagnosed as schizophrenic by the psychiatric staffs of their respective hospitals.

An attempt

was made to secure relatively acute (i.e. ”undeteriorated”), coopera­ tive cases. None of the patients, at the time of testing, was being subjected to either physical or chemical therapy. Statistical data pertaining to the age, educational level, duration of hospitalization* number of hospital admissions and diagnostic sub-type composition of the SX and SC groups are presented in Appendix A.

Examination of

these data indicates that the group means, medians and ranges for these variables were substantially similar, with th® exception that the SC group contained a somewhat higher proportion of paranoid

11 schizophrenics than did the SX group*

An indication of the success

achieved in obtaining acute cases for th© experiment is the relatively brief period of hospitalisation of the SX and SC patients (the group means are 1.2 and 1.1 years respectively). Normal Groups The subjects comprising th© NX group war© largely semi­ skilled workers recruited through the personnel office of a large neighboring manufacturing concern; th® NO group consisted primarily of city firemen.

Many of these normal subjects had been previous

participants in research projects at the Worcester State Hospital. Age and educational data for the two normal groups are also presented in Appendix A. Both of these groups were substantially similar in terms of age and education, although th® MX group had a somewhat greater age range.

Comparison of all four subject groups for age and

educational level indicates little difference on the average between the schizophrenic and normal subjects. Apparatus The experiment was run in a sound-shielded room at Wor­ cester; testing at the Grafton State Hospital was done in a room at the end of a quiet corridor removed from the patients1 ward.

Subjects

were seated at a table, with E seated on the opposite side behind a celotex screen. The stimuli used were tones presented by means of a batteryoperated General Radio beat frequency oscillator (Type 613-B), hooked

i2 vp through a matching transformer to a deasea $*• M. 100 Electrodynamic speaker. This oscillator is ths sort employed by Wendt {£$) ia determining the auditory acuity of monkeys and by Borland (?) in his studios of stimnlas generalisation • Tones of one-second duration were presented automatically every Xk second# by use of a Telechron motor-driven mechanical timer. A small switch lever set ia a wooden her m the table in front of the subject could be manipulated in either a f©reward (pull} or backward (push) direction to step a Springfield Interval Timer which started automatically with the presentation of a tone. A flashing red li^t on 3B's side of the screen indicated the direction of the subject *8 finger movement. At the side of the screen, facing the subject, was a white cardboard scoreboard in a black picture frame 16 3/4 laches by 15 l/4 Inches. A red column in the center (1% 3f k inches by Z XfZ inches) mode it resemble in miniature the familiar Community Chest type of scoreboard.

A cutout gold cardboard was superimposed on the

red column and could be moved up and down over it to any extent de­ sired by the experimenter. A gold-painted oardboard box, 3 inches by 3 1f k inches, was set on the subject's side of the screen at the base of the score­ board.

Its ground glass front (1 1fk inches by 3 3/4 inches) covered

a cardboard cutout reading 'Bight' which became visible when the box was illuminated. A red cardboard box of the same dimensions was added during the Beward-Punlsfcment condition. when it was illuminated.

This box read 'Wrong'

The position of a pre-selector switch

IK operated by S determined whether the subject’s response would light the •Right* box, the *Wrong* box (when it was present) or neither. In addition to the scoreboard and the signal boxes a bar of candy and five cigarettes were also placed on the table in front of the subject. A photograph of the apparatus as seen from the subject* s position is presented in Figure 1. Figure 2 shews the wiring diagram for the experimental set-up. procedure Saeh subject was run on two successive days.

The first

day’s session served as a preliminary to the experimental sessions and had a two-fold purposes

To permit the subject to become ac­

quainted with B; and to enable S to observe the subject's behavior on a simple task similar to that presented in the experiment proper, in order that S might eliminate any subjects who appeared incapable of participating in the experiment. This session was identical for all four groups of subjects. Prior to the presentation of the practice task E interviewed each subject for approximately 1$ minutes, during which the subject m s told that he was to participate in a test designed to measure his reaction time and ability to remember tones.

Infomat ion with res­

pect to age, occupational and educational status and evidence of hearing loss ox*

1ity* Weis obtained.

In Stdditioni,

svt*

jects were questioned regarding any history of previous mental break­ down.

This interview was, in E*3 judgment, fairly effective in put­

ting most of the subjects at ease and enlisting their cooperation#

14

Figure 1.

Subject*8 View of the Apparatus. (Switch Lever in Foreground)

JO;D3'Puf

r.^Ds/’ dfny

5

*

^ s

> ^

Q

^

2.31

2.1|6

2.11

2.2U

2*65

f^O

4

’ '

10

9

FREQUENCY

OF PULL RESPONSES

8

7 6

5 4

MEAN

3 • NX-P N=24

2

- - - # SX-P N=24

1

s0 B’

A’

C*

D’

E*

STIMULI

figure k .

frequency of Pull Besponsee Ub Each 3tianiXni by Schisophrenic and Hemal Subjects Under iteward-Funishraent Condition*

Mean

r?i Inspectlou of the curves in Figure

reveals a greater dis»

parity in the shapes of the curves for the two groups than was obtained under lew&rd Only, particularly at Points G f, $*s and SS1. When the mean frequency of total responses to the four generalised stimuli was computed for both groups, it was found that the normal subjects had a mean of 13*96 and a S.JB. of 5*03» whereas the mean and S.3D, for the schisophrenic* were 18.21 and J#§3 respectively.

The jfe for the dif­

ference between these means is 2.25, which is significant below the

5$ level for *46 degree® of freedom.

It would thus appear that under

this condition the schizophrenics showed a reliably greater tendency to respond to the generalised stimuli as though they were equivalent to the training stimulus.than did a comparable group of normal subjects. The results under this condition also show a marked decrement for the S2&»P group in the mean frequency of pull responses to the training stimulus as compared with the M*»F group.

A Jr-best of the

mean difference between these two groups at ton® B 1produces a ratio of 2.3S# which is significant at the &*5# level, for *46 degrees of freedom. Since the normal subjects responded more frequently to the training stimulus and lose frequently to the generalised stimuli than did the schizophrenic subjects, the difference between the two groups In the steepness of the response gradients from B 1 to B1 ml^t be expected to be highly reliable.

A G-test of this hypothesis yielded

a aowsal~&eviate value of 3.61, which is significant at the .0*4$ level. It could be concluded with great confidence, therefore, that the

schisophrenic* Showed a markedly flatter response gradient than did normals under the Bsward-Funishment condition. As the preceding discussion indicates, the schisophrenic subjects gave significantly fever correct responses to the training stimulus than did the normal subjects.

It might therefore be argued

that the large differences between the B 1 to B1 gradients of the two groups under the JSew&rd-Punishment condition are attributable to the different amounts of reinforcement experienced by the two groups at B 1. To investigate this possibility, a separate comparison was made of the responses to the generalised stimuli of those schisophrenic #

and normal subjects who could be matched on the basis of their frequency of response to the training stimulus on the test trials*

It was pos­

sible to accomplish an exact matching of 15 schisophrenic and 15 nor­ mal subjects on this basis. Comparison of these two selected groups under Seward Only indicated very little difference between their em­ pirical response gradients under this condition*

There were marked

differences between them, however, under the Howard-Puni shment con­ dition* The relevant data for these selected subjects under HowardPunishment are presented in Table IV end plotted graphically in figure 5* from these data it may be seen that althou^i the scores of these selected schisophrenic and normal subjects were identical at th. training stimulus (B'), tin fr.qu.ney of rcepons. to tin g.nsralUsd stimuli was much higher for the schisophrenic subjects.

A j^-test of

Tabl« X? Means and S.D. *8 of Distributions of Fall Responses To iS&ch Stimulus by ftfistuhed SX-F and H3HP Groups (l for each grovep * 1$) Group

Matched N5HP

Matched SX*F

Stimuli A'

£'

£'

£'

M

6.07

7.33

3.93

3.00

1.27

se

2,*t6

1.81

1.98

1.79

2«lS

u

6-73

7.33

5-73

fc.13

3-«7

SD

1.98

1.81

2 .lU

£.19

2*87

34

10

9

MEAN

FREQUENCY

OF PULL

RESPONSES

8

7

6 5 4 3

2

•- --------- #

MATCHED NX-P N=15

•---------- •

MATCHED SX-P M=15

1

0 A’

B

C’

D1

E’

STIMULI

Figure 5* Mean Frequency of Full Besponaee To Sack Stimulus ly Matched Schisophrenic and Normal Subject* Under Bewar&^Puaishrnent Condition.

10

9

MEAN FREQUENCY

OF PULL RESPONSES

8

7

6 5 4 3

2

#

MATCHED NX-P N=15

— •

MATCHED SX-P N=15

1

0 A’

Bt

C'

Df

E'

STIMULI

Figure 5* Wean Frequency of Pull Beeponeee Sfe s&ch Stitnulu* By Matched Sehiaephreaio and Eorraal Subjects tinder fieward-’Punishroeni Condition,

85 th, dtff,r»nc, between the frequency of pull roopoaooo to tho four generalized stimuli for thooo seleeted groups {KX “ B ~ 1*U27»



H.«5j SX • U 8 80.U7, 3.B. S 7.80) yields a value of 8.59, *Mch is significant at the 2* 5# level. Using the differences in the frequency of pull responses to B* and I!9 as a measure of the steepness of the response^ gradients for the two seleeted groups, a normal deviate value of 2.2$ obtained by means of the latest indicated that the differences between the response gradients of these two groups were significant at the 3# le v e ls

When a £ for the difference between the means of the B 1 minus

B* scores for the two groups (IS «* II 3 6.07? SX - II ® 3»^7) computed, * the difference proved significant at the 1ft level ** Bespits the skewness in the individual group distributions of ijfUg1 scores, the use of t, with these matched measures was not pre­ cluded since the distribution of difference scores was essentially normal* I . . . - — ............ ..I .

.....................

(t « 2.99, I k d.f.)«

n il

r

■ n

These results lend strong support to the con­

clusion that the differences between the obtained response gradients for the schizophrenia and normal subjects under Beward-Punishment was, at least in part, independent of differences betwben the two groups in the number of reinforcements received at the training stimulus* Beward Only vs, Bsward-Ponlafamcnt The mean frequency of pull responses to each stimulus under the Beward Only and Beward-Funishment conditions have been re-plotted

%

in Figure 6 in such a way as to facilitate comparison of the effects of th© two conditions upon th© performance of both normal and schiso­ phrenic subjects.

Inspection of the curves for the normal subjects,

indicates that the introduction of the Bsward-Funishment condition had a marked effect on the ability of these subjects to differentiate among the stimuli* fhe improvement manifested under this condition is indicated by the results of a J^-test of the mean difference in the number of responses to the four generalized stimuli by this group under the two conditions.

For

the mean was 19*79 and the S.D.

was 6.38, while the same individuals under Beward-Pvrniehment (HXrP) had a mean of 13*9& and a S.S. of 5*03* ®i® obtained jfe of 3*9^ considerably below the .1$ level of significance (£3 d.f*). Under fteward^Ptmishmenfc, the differential responsiveness to the training stimulus and to the farthest generalised stimulus was also sharply increased in th© normal subjects.

A U test of the

difference between th© distributions of 1-1 scores (Beward Only) and I 1-!1 scores (Beward-Punisbment) yields a value of 2*58, significant at 1$. This clearly established the superiority of the BewardPunishment condition for normal subjects in facilitating their dif­ ferentiation between the training stimulus and the farthest generalized stimulus* By contrast, the curves for the schizophrenic subjects run under both Beward Only and Beward-Puniehment present a rather dif­ ferent picture* As in the case of the normals, the schisophrenic subjects

MEAN

FREQUENCY

OF PULL

RESPONSES

37



SX-R N=24

O

NX-R H =24

- - •

SX-P N=24



tiX-P N=24

0 --

STIMULI

figure 6. Mean Frequency of Pull Besponses fo Bach Stimulus By Schisophrenic and Hdnaal Subjects Under Beward Only and BewarA-Punishraent Conditions,

38 tended to respond lees often to the generalised stimuli under BewsrdPuaiehmenl than under Beward Only.

3Ehe mean and S.23. of the number

of pull responses to the four generalised stimuli by the SX-B subjects under Beward Only were, respectively 21.1? and 6*72- *! Bies© same subjects under Beward-Punisfament (S3M?) had a mean and S.D. respectively of 10.21 and 7*f3 * ®h© & for the .difference btittaesa the means (matched measures) was S.tyi, which was significant at the 2«3$ level of con­ fidence for 23 degrees of freedom*

It could be concluded therefore,

that.the schisophrenic grouiymad© fewer prohibited responses to the generalised stimuli under Bcward^.Punishment than under conditions of Inward Only* Whereas the normal subjects, however, showed increased re» sponsivenes© to the training stimulus under the Beward~Pimishment condition as compared with Beward Only, the schisophrenic subjects made fewer pull responses to the training stimulus during the Beward* Punishment condition*

Although a £*test (for related measures) of

the difference between the mean number of pulls at tone B (training stimulus under Beward Only) and tone B* (training stimulus under Eeward-Punishment) was significant at only the 7$ level (t a 1*90* d*f* s 23), it does not appear improbable, that this decline in re* sponslveness to the training stimulus represented a genuine phenomenon* In any event, the decline in the number of incorrect responses to the generalised stimuli by the schizophrenic subjects under Beward-Pmiiehment clearly gave no grounds for supposing that they differentiated more successfully between the training stimulus and generalised

stimuli than they had done under Beward Only*

When the B to S gradient

under Beward Only w&e compared with the 1* to $* gradient under Beward* Punishment for the schisophrenic©, a tf-test yielded a normal-deviate value of *10* On this hasis it could he concluded that the schizophrenics unlike the normal subjects, showed no increase in their ability to differentiate between training end generalised stimuli due to the Introduction of punishment*

Tm sole effect of punishment it appeared,

was to decrease responsiveness to the generalised stimuli and per* haps, to the training stimulus* Beward Only ** Session & Xt will be recalled that schisophrenic and normal control groups were run under a second Beward Only condition in order to terrain® whether the results for the Bsward-Fanishment condition might reasonably be ascribed to factors such as learning, fatigue or other progressive effects which might enter as the subjects continued to practice the task*

If such factors were responsible for the effects

observed under Beward**Funishment, it mi#it be expected that continued practice aider Beward Only would also tend to produce them*

The re*

suits obtained for the schisophrenic and normal controls are presented in Table If, which gives the mean and S.D» of the pull response fre* quencies to each stimulus during the second Beward Only session* Figure 7 shows a graph of the mean frequencies at each point for the two groups under this condition*

40

3febl» f Means and S.D* *fi of Distributions of Pull Besponaos To Inch Stimulus for SQ*M and Group© (IS for each group £ 16) Stimuli

Group

sss-m

B'

I'

5.88

^•3«

3.75

2.71

2.1S

2.U7

2. ®

7.00

?.6f

6.19

g.W»

U.go

2.12

l.?6

2.28

2.83

2.98

A*

1'

M

6.13

7.3s

SB

3.*A

M SB

41

10

RESPONSES

6

MEAN

FREQUENCY

8

OF PULL

9

7

5 4 3

#

NC-R2 N=16

— #

SC-R2 N=16

2 1

0 B'

Cf

D'

E*

STIMULI

Figure 7*

Mean Frequency of Full Besponses U?o £Jach Stimulus By Schisophrenic and Normal Control Subjects, Under Second Beward Only Condition.

As ill the initial Beward Qnly condition, these results indicate that the SO-UB group produced a higher mean frequency of pull responses to all four generalised stimuli than did the KG-BB subjects.

A* be»-

fore, however, a ,1,-test of the difference between the mean (SO-$2, M « 23*13* S*S. m 6*96; 8S-BB, U * 20.13,

3 7-12) frequency of

responses to the four generalised stimuli was not significant (t ■ 1*15# F s 20-3®^ f©r 15

might be expected, the difference in the

1* to B* gradients for the two groups wee also insignfieant, the U**teet yielding a noreal deviate value of *57* indicating a difference that would be exceeded by chance more than 90$ of the time*

In view of the

absence of any notable differences between the SC and HO groups under the second Beward Only condition, it does not seem reasonable to attribute the striking differences between the SX«P and HX*P groups to practice alone* fhis conclusion Is supported by direct comparison of the second Baward Only condition with the Bcward-Funisfement condition*

In

terms of responsiveness to the generalised stimuli and in terms of differential responsiveness to the training stimulus and the farthest generalised stimulus, the relationships between the second Beward Only condition ^and the fieward^Funishment condition were in every Instance essentially similar to those found between the first Beward Only con­ dition and Beward-Pualehment. IShere was no reason, therefore, to suppose that there was greater similarity between the Beward-Funishment condition and the second Beward Only condition, as a function of practice*

4n Comparison of yirst and Second Sessions. Beward Only Although it seems clear that continued practice under Seward Only did not have th© differential effect upon the behavior of the schizophrenic and normal subjects found for the Beward“Punishraent con­ dition, it was still possible to ask whether continued practice under the the condition presented initially affected the behavior of the two groups at ail.

In order to facilitate comparisons of the perfor­

mance of the control groups on the two sessions of Beward Only, the mean frequencies of pull responses to each stimulus by the 3C and 2& groups during both sessions are presented in Figure S.

It may be

seen that th® curses for th® two sessions are very similar for both sehlsophreaies and normals.

There appeared to be no increase in the

differentiating ability of either group due to practice.

Indeed

there was a very slight tendency for both groups to make more pull responses to the generalised stimuli on the second session than on the first*

The gradients from B to 13 and B* to B*, however, were

practically identical.

In neither case were the obtained differences

between these gradients greater than would be expected on the basis of chance alone 80% of the time, as determined by t? tests.

On the

basis of these results, there seems little reason to beliewe that continued practice following the first session of Beward Only had any masked effect on the performance of either group.

44

eoaMMMwws.r • j ssosl

m

MEAN

FREQUENCY

OF

PULL

RESPO N SES

10 r

A

• ------------ •

S C -R 1 H = 1 6

•- - - - - - • NC-R1 N=16

• ----------- ♦

S C -R 2 N = 1 6

•--- • NC-R2 N=16

B

C

D

E

A

B

C

D

STIMULI

W ® 8*» 8 . Iteaa yrttpasncy of fall SespoaaM So £ach Stloalaa By Sehliopfcjreale and Soiaal Coatcal Sabjacta On&ar Xa» a**urd