THE mm or mwas ON
com-515mm BEHAVKDR o;
scwzovmwc AND NORMAL mums. ,

mmmmnumamm'
(1912th STATE COW
Nsé Papania
1953

 

:5 This it to certify that the ‘
thesis entitled f,

2

I

THE EFFECTS OF FAILURE (N GOAL—SEMI!“
BEHAVIOR OF SCHIZOPHRENIC AN'D NORMAL SUBJECTS.

presented by t

has been accepted towards fulfillment
of the requirements for

.; flaw “M
I ' Major professor
Date W/

 

      

 

 

THE EFFECTS OF FAILURE ON GOAL—SETTING- BEHAVIOR

OF SCHIZOPHRENIC AND NORMAL SUBJECTS

By

Ned Papania

A THESIS

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements

for the degree of
DOCTOR OF PHILOSOPHY

Department of Psychology

1953

Ii'qfibl 9

ACKNOWLEDGMENT

The writer wishes to express his sincere thanks to Dr. Albert 1.
Rabin, his committee chairman, and to Dr. David Pearl, Assistant Chief
Psychologist at Fort Custer Veterans Administration Hospital, ihose
suggestions, encouragement and cooperation contributed so significantly
to the completion of this dissertation.

He is also greatly indebted to Dr. Milton Rokeach and Dr. Donald

M. Johnson for their invaluable assistance.

11

31.1285)

ul‘ll All! lit...

THE EFFECTS OF FAILURE ON GOAL-SETTING BEHAVIOR

OF SCHIZOPHRENIC AND NORMAL SUBJECTS

3!

Ned Papania

AN ABSTRACT

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

DOCTOR 0F PHILOSOiHY
Department of Psychology

Year 1953

Approved 4ég%>4éZP57zt36%Z¢<E:;<%£:;"’——_——‘_’

Ned Papania
The study concerned itself with the post failure goal setting be-

havior of a group of 32 mixed schizophrenics and a group of 32 “normal”
subjects equated for age, education, intelligence and initial performance
on the tasks to be used.

A level of aspiration setting was utilized for two tasks: (a)
solution of simple addition problems (b) tweeser dexterity. These tasks
were presented to the subjects as tests of intelligence and mechanical
ability in order to 'ego involve” them.

The two major groups were divided into two equated subgroups of 15
subjects. All four subgroups were given five practise trials on each
task and five success trials during ehich they achieved their levels of
aspiration.

From.this point on.the eXperiment consisted of a failure trial. in
which level of aspiration was not achieved, followed by two success
trials (referred to as post failure trials). This sequence was repeated
five times for each task; The failure trials vere differentiated into
strong failure (in which only 50 per cent of the goal eas reached) and
Mild failure (in which 90 per cent of the goal was reached).

Sixteen schisophrenic subjects and sixteen controls were exposed
to strong failure, an equal number being exposed to mild failure.

Sin variables were studied, vis: level of aspiration, discrepancy
score. efficiency. reaction time, errors, and responses to questions
asking for their subjective feelings after failure and success.

The hypotheses stated that after failure both schizophrenic groups

should give higher levels of aspiration, higher discrepancy scores and

‘vfi

Ned Papania

longer reaction times than the controls. Further that they should she!
lorer efficiency, greater error scores than the controls, and finally
that they should attempt to deny subjective feelings of failure more
frequently than the controls. Also. that the schizophrenics in strong
failure should show these reactions to a greater extent than those in
mild failure.

The data, analysed by means of the one tailed t test. substantiate
all of the hypotheses. The data further indicate that there is some
behavioral disruption due to failure in the control strong failure
group. Though they do not react as strongly as the schizophrenic groups,
they do shoe a higher level of aspiration after failure than they did
in success. They also shoe lover efficiency and greater reaction time.
The above findings also apply when they are compared to the control
mild failure group.

The findings are interpreted in accordance with Cameron's theory

of reaction sensitivity to failure.

TABLE OF CONTENTS

ACKNOWLEDGMENT . . . . . . . . . . . . . . . . .
LIST OF TABLES . . . . . . . . . . . . . . . . .
INTRODUCTION . . . . . . . . . . . . . . . . . .
REVIEW OF THE LITERATURE . . . . . . . . . . . .

Early Studies in Level of Aspiration . . .

Special Characteristics of Level of Aspiration Scores

Individual Differences . . . . . . . .

Reliability of level of aspiration scores . . .
Generality of level of aspiration behavior . . .

Levels of reality . . . . . . . . . .
Forces Operating in Goal Setting . . . . .

General forces . . . . . . . . . . . .

Ego involvements and levels of aspiration .

Characteristics of the task . . . . .
External reference scales . . . . . .
Internalized reference scales . . . .

Attempts to Use Goal Setting Behavior Diagnostically

Level of aspiration and traits . . . .
Studies With Clinical Groups . . . . . . .
Studies with neurotics . . . . . . . .

Studies with Psychotic groups
Summary . . . . . . . . . . . . . . .

SCHIZOPHRENIA, FAILURE AND FRUSTRATION . . . . .

Theories of Schizophrenia . . . . . . . . .

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

The Effects of Frustration on Goal Directed

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

Hypotheses . . . . . . . . . . . . . . . .

Behavior.

0 O O O O

MU'IKJ'I

OKOkD on

12
13

16
18

19
19
22
22
23
31
31
33
31+
_ 37
3s

METHomLO GY O O O O O O O O 0 O O O O O O O O O O 0
Requirements of The Experiment . . . . . . . .

Requirements in the selection of groups .
Requirements in the selection of the task

The Groups . . . . . . . . . . . . . . . . . .
Basis of choice of the control groups . .
Basis of choice of the schizophrenic group

The normative data . . . . . . . . . . . .

The TaSkS e e e e e e e see e e e e e e e e e e
The Arithmetic problems . . . . . . . . .
The second task . . . . . . . . ... . . .
Time limits e e e e e e e e e e e eseee e

Experimental Procedure e e e e e e e e e e e 0

General set . . .

Instruction e e e s s e e e e e e e e e e
General procedure eeeeeeeeeeee
TheC.A.S.................
The practice sessions . . . . . . . . . .
Thesubgroups ..............
The success trials . . . . . . . . . . . .
The success—failure series . . . . . . . .
Timing..................
Order of Presentation . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . .

FINDINGS . . . . . . . . . . . . . . . . . . ... . .
Method of Analysis . . . - . . . . . . . . . .
Results on Level of Aspiration . . . . . . . .
Results on Discrepancy Score . . . . . . . . .
Results on Efficiency Score . . . . . . . . . .
Results on Number of Errors . . . . . . . . . .
Results on Reaction Time . . . . . . . . . . .

Results on Questions . . . . . . . . . . . . .

iv

0 I O O 0 O O O

Page

DISCUSSION . . .

Comparison to Other Schizophrenic Groups . .

Comparison Between Schizophrenics and Controls

Implications

SUMMARY AND CONCLUSIOHS

BIBLIOGRAPHY . .

Page
82
82
83
85
87
89

......

......

......

......

oooooo

 

Table

it!

5a

5b

LIST OF TABLES

A Comparison of the Groups with Respect to Some normative
Factor. C O C C O C Q 0 O I O O O O O O ‘ O O O O I O O 0

Procedural Sequence of Tasks for the Experimental Groups.
”can Level of Aspiration for No Schizophrenic and Two

Control Groups in Conditions of Success and Failure on
Two'fillho.......oo.......o.....o

.Differences Between Types of Trials on llenn Level of

Aspiration for he Schisophreuic and he Control Groups
in Conditions of Success and Failure on Two Tasks . . . .

Differences in lean Level of Aspiration for he Schiso.
rhrenic and he Control Groups in Conditions of Success
er11WCOHMTQmeeeeeeeeeeeeeeee

flean Discrepancy. Score for he Schisophrenic and fee
Control Groups in Success and Poet-Failure Trials on
m “M O O O O O O O O C C I O O 0 O O O O O O O O C 0

Differences Ietveen Types of Trials on Kean Discrepancy
Scores for Two Schieophrenic and the Control Groups in
Conditions of success and Failure on he run . . . . .

Differences in than Discrepancy Scores for he Schles-
phrenic and Two Control Groups in Conditions of Success
“FAH‘NOITIOTIJEIo.....o.o....o.o

Mean Ifticiescy Score for he Schisophrenic and he
Control Groups in Success and Post-rsilure i'riele es
N?BMeeeeeeeeeeeeeeeeeeeeeeee

Differences Between Types of Trials on wean Efficiency
Score for he Sohiscphrenic and he Control Groups in
Conditions of Success and failure on he Tests . . . . .

Differences in non Efficiency Scores for he Schisc-
rhrenic and Two Control Groups in Conditions of Success
MMImOI‘l‘NTa'kIo......o........

lean hunter of Errors fer be Schizophrenic and i've
Control Groups is Success and Post-Jailure i'riale on
on. 7‘“ C O O O O O O O O O 0 O O 0 e O O O O O O O O 0

P84:-

1"5
56

60

61

65

70

72

Table

1a

Differences Between Types of Trials on loan number of
Errors for he Schisophrenic and Two Control Groups in
Conditions of Success and Failure on One Task . . . . .

Differences in Ilene number of Errors for fee Schin-

phrenic and Two Control Groups in Conditions of Sacco"
”drailu.onou‘r8aeeeeeeeeeeeseeee

Mean Reaction Tine ( in seconds) for Two Schizophrenic
and Two Control Groups in Success and Post-.l‘cilure
Trialeon'l'vo‘raeke..................

Differences Between Types of Trials on “can Reaction
IPine for Two Schisopbrenic and Tm Control Groups in

Conditions of Success and failure on Tvo Tasks. . . . .
Differences in flean Reaction Time for Two Schizophrenic

and Two Control Groups in Conditions of Success and

Failure on Tue Tasks

O O C O O O O I O O O O I O O O O

'11

73

1h

77

79

.ll'll'll l Ill (I [ll I'll I [{‘I [II (I' V

 

INTRODUCTION

KnoWIedge of the goal a person sets for himself is important for
the true appraisal of the efficacy of his behavior and for the under-
standing of his reactions to environmental stress. Such goals may be
explicit and formulated by the individual himself who announces, for
instance, his intention of running a certain distance in a given time,
or the formulation may come from the environment which prescribes a
standard of achievement ihich the individual has to attain. In either
case the setting of the goal is the first step toward.action ehich.will
culminate in attainment or nonpattainment, in success or failure. The
subjective feeling of success or failure will in its turn influence
other goal settings, as well as other goal oriented behaviors.

A comparison of the goals of different individuals as well as a
comparison of their reactions to success and failure may thus prove a
‘useful approach to the study of personality makedup.

The present investigation is concerned with the goal setting be-
havior of schizophrenic subjects in a situation which threatens than
with failure in various degrees. To an extent the writer hopes to
show the connection between failure situations and the “unrealistic“
behavior toward the world shich is clinically noted in schizophrenia.

Specifically the writer wishes to correlate the changes in goal

setting behavior with changes in the outcome of preceding experiences,

with previous success or failure.

2

However it is difficult if not impossible to compare directly the
principal goals in life of am individual or of any group of individuals.
The problem has to be narrowed down to enable objective measurement.

One of the better known approaches to this problem has been via the
studies in level of aspiration. Here one can observe the setting of
goals, their accomplishment or non-.accomplishment and the effect of this
success or failure on future goal setting, and on objective performance
in certain specified tasks. .

As will be seen in the review of the literature, these studies be-
gan with qualitative observations, and only recently has there been
some attempt to quantify the approach. Though there is nothing sacred
in quantification per se, it is of some help particularly when one is
still in terra incognita.

In the writer's opinion, Frank's (11$) introduction of the discrep-
ancy score (the difference between previous performance and subsequent
goal), lent clarity to the further studies of level of aspiration. It
represented an advance over Hoppe's (39) qualitative method of evalua-
tion. Rotter's pattern analysis (62) represents an even more fruitful
way of dealing with the problem. 4

Measures of rigidity or lack of responsiveness to success and failure,
as well as objective measures of performance in success and failure con-
ditions also have been of help in clarifying the complexities of the
individual's reaction to success and failure in a goal setting situation.

The control of success and failure, first introduced by Jucknat (’45),

again added to the certainty of measurement. While the studies of

3
Bayton and Holt (M, 36) which have attempted to control ego involvement
have brought further facets of the problem to light.

Recently Steisel and Cohen (67) have met another problem in the
methodolog of level of aspiration studies. In their study they de-
fined failure in terms of two criteria: (a) non-attainment of level
of aspiration (b) social disapproval. Their method allows them to set
up degrees of failure beforehand and to control for them. Though some
objections may be leveled at this equation of subjective feeling of
failure and objectively controlled failure, it must be admitted that
as yet no accurate method of ascertaining subjective feelings of
failure exists, and that the controlled method, by defining failure
independently of the results, does add to precision of measurement.

Thus with the passage of time level of aspiration studies have
become more precise. Their use in studies with certain clinical groups
which was at first decried because the method itself had not been suf-
ficiently explored, seems at present advisable, and perhaps even neces-
sary; for it seems to the writer that success and failure and the re-
actions of persons to them, in both performance and setting of future
goals is a most important area in the field of psychopathology. Schizo-
phrenia, as one of the most prevalent and least understood of the men-
tal disorders, bears definite investigation along these lines.

Thus, it my be asked: Does the schizophrenic set his aspirations
beyond his. performance ability in a greater measure than do other
people? What are his reactions to mild and strong failure in settings
of success? How do his reactions to mild and strong failure differ

from that of nomals in a setting of basic success? Is the schizophrenic

able to modify his aspirations in line with his actual performance as
much as others? Does the pattern of reaction to success and failure
seen in schizophrenics resemble that of those who have had.many fail-
ures in life?

These are the basic questions which this investigation attempts
to answer, in.the hope that some small particle of knowledge'may be

added to the existing concepts of schizophrenia.

u

l.[[[‘lllll'l[ll[lllll(f

REVIEW OF THE LITERATURE
Early Studies ig_Level g£_Aspiration

The tenm "level of aspiration“ first came to the attention of the
psychological world as a result of Tamara Dembo's study "Anger as a
Dynamic Problemt' (6). By presenting her subjects with insolvable prob-
lems she was able to elicit reactions of anger directed against than-
selves. Dembo accounts for this via the construct of “level of aspira—
tion,‘I ihich she felt determined whether a given performance would be
acknowledged as successful or not by the subject.

A second student of Lewin's, F. Hoppe (39), continued to investi-
gate this phenomenon described by Dembo. By observing his subjects in
a wide variety of simple tasks given in a relatively free situation, he
felt that he was able to infer experiences of success and failure from
their spontaneous utterances and general behavior. These subjective
experiences of success and failure seemed to depend, according to Hoppe,
not as much on the actual performance, as upon what the individual 61b
pected of himself.

Thus, if in a task of target shooting, the subject feels fairly
certain that he can hit the fifth ring from the center and then only
hits the seventh, he is likely to express his disappointment over his
experienced failure. However, if in the course of subsequent trials
he continued to hit within the seventh and eighth rings from the center,
a hit on the sixth ring would be experienced as success, as he no longer

expected to hit any closer than the seventh ring. In this case the

subject experienced failure over a particular goal at one time, and
success over a lesser goal at another time. This specific goal, the
level of difficulty which a person chooses to attempt in a situation
where he has a choice between tasks of different degrees of difficulty,
is what Hoppe defined as "level of aspiration.“

The “level of aspiration,” as can be noted from the above is not
static and absolute, but tends to fluctuate with the fortunes of the day.
. Hoppe found that success experiences tended to raise the level of aspi-
ration while experiences of failure lowered it. However, this generali-
sation is only an approximate one, as there seem to be many individual
differences in goal setting. It was the existence of these differences
which led Hoppe to stress the value of this technique in the assessment
of personality.

Of course the difference in "level of aspiration” is not only a
function of individual differences, but also of the task.itself. Hoppe
describes a "bordersone of ability,‘ that is a certain range of task
difficulty within which this phenomenon seems to occur. Thus tasks
perceived as too simple or too difficult do not elicit levels of aspira-
tion.

Hoppe's work may be seen as an extension of Dembo's on a qualitative
plans. It was totally dependent on the acuteness of observation and
liable to the errors of all purely qualitative evaluations.

Jucknat (#5) whose work followed Hoppe's, undertook to quantify the
method. -She used.a series of paper and pencil mazes of increasing size
and complexity. These she arranged on a table in the form of a ladder,

with the smallest and easiest on the bottoms

The subject was informed of the difference in the difficulty of
the tasks, asked to choose any one of the mazes and instructed to com-
plete it at another table. The particular mass which the subject chose
was considered his momentary level of aspiration, as it was the one
which he apparently expected to do. The raising or lowering of the
level of aspiration could be judged by his subsequent choices.

Jucknat also introduced the technique of controlling for success
and failure. This was accomplished by introducing a set of unsolvable
mazes which were given along with the ordinary set. She arranged her
experiment so that one group of subjects first experienced failure and
then success, while another group experienced success first. Jucknat's
quantifiable study substantiated Hoppe's qualitative pioneer works

One of the most persistent investigators in this area has been
J. 1). Frank (in, 15, 16, 17, 18, 19). He has used tasks of a somewhat
different nature, such as printing with the fastest possible speed, and
ring tossing.

In.Frank's studies the subjects are first given a few trial runs
to acquaint them with the task. ‘At the end of each of these practice
trials they are told.how long they have taken to do it. After a number
of practice trials they are then asked to estimate the time they will
take to complete the task on the nexttrial. .At the end of this trial
they are again informed of their scores and again asked to state their
next aspiration. Ry this method the goal is expressed in the same units
in which goodness of performance is measured, (i.e. time) thus enabling
direct comparison to be made between level of aspiration and level of

performance.

8

-Frank'uses two principal scores to evaluate goal setting behavior
(a) the goal discrepancy score (D score) which is the difference be-
tween previous performance and subsequent goal and (b) the percentage
score of typical reactions to success and failure (i.e. lowering level

of aspiration after failure and raising it after success).
Special Characteristics 9£LIevel g£.Aspir§tion Scores

Individual differences. Frank's primary interest is in the psycho-
logical determinants of the level of aspiration and in the individual
differences observed in this situation. He assumed past performance
at any thme depends primarily on the relative strength of the following
needs: (a) the need to keep the level of aspiration as high as possible
regardless of level of performance, (b) the need to make the level of
aspiration approximate the level of future promise as closely as pos-
sible, (c) the need to avoid failure where failure is defined as a level
of performance below level of aspiration regardless of its absolute
goodness.

He also assumed that certain types of behavior on level of aspira-
tion express what are usually termed personality traits.

Frank.offers evidence to show that the difference between level of
aspiration and level of past performance in a given task represents a
relatively permanent characteristic of personality, and that this per.
manence can be demonstrated regardless of the type of ability that the
task.requires. He presents data to show that the traits manifested in
the experiment are to a large extent independent of the physical nature
of the tasks. Frank suggests that the size of the average level of

past performance is due to the involvement of the "ego—level“ of the

9

individual in the task, as shown by self-competition and social pres-

sure.

Reliability of l_e__v_g_l_ of aspiration scores. Many subsequent investiga-
tions of the level of aspiration have been carried out in the last
decade. These include investigations of the reliability of the technique
by means of the split half reliability method and by retests after an
interval.

Frank (1)4) found reliability coefficients ranging from 0.57 to 0.75.
the lowest correlation (test-retest) obtaining in quoit tossing which
he explains as a function of the fact of the task's, game like, and thus
irreal character. These findings were challenged by Gould (26) who
found reliability to be quite a bit lower. Gardner (2%) using controlled
success and failure as well as various tasks obtained reliability co-
efficients ranging from 0.79 to 0.98. Heathers (32) also found that
“as in other aspiration studies the aspiration scores were found to be
highly reliable. The corrected odd-even correlation between the mean
deviation scores ranged from 0.93 to 0.99 for all tasks with every group
of subjects.” (p. 3348).

Lewin .e_t_ 11‘. (51) state “The difference between the results of Gould
and frank is perhaps this to the greater diversity of the tasks used by
Gould and the, lesser amount of control over the subjects attitudes in

the Gould experiment." (p. 3147).

Generalitz of; level of aspiration behavior. Generality of the behavior
has also been examined by several psychologists (11+, 21+, 26, 32). The

correlations were all positive and such as to warrant Frank's description

10
of the behavior as a I'relatively permanent characteristic of personality."
(1n). This generality is not absolute and is affected to varying de-
grees by situational factors, by similarities of abilities tested as

well as by the amount of success and failure experienced.

nggl§.g£_realitz. Another factor which has been examined and which
also affects generality is the level of realityairreality.

Gould (26) found that the level of aspiration statement had four
possible meanings (a) the average level or the actual score expected
the next time (b) the least level or the lowest score expected (c)
greatest or highest score expected (d) a.mixed grouping in which the
subjects shifted the three levels according to the momentary situation.
She found that different persons used different levels of approach to
the request for an aspiration level as a function of their defensive
needs. Thus some consciously gave a lower level than they expected to
get, while others gave a.higher level than they could hope to reach.
She concludes that "the evidence points to the fact that the estimates
given in this and similar situations do not accurately express what
the individual is striving for.“ (p. M7).

Frank (17) experimentally varying these levels found that sub-
jects asked ”what do you think you will do“ were more realistic than
those asked “that do you intend to do."

Festinger (13) found that a group asked Iwhat score would you like
to get ”has significantly higher discrepancy scores than a group asked
“what do you expect to get;' the former group‘also evidenced greater
variability in levels of aspiration. These results were corroborated

by Irwin and Mintzer (42) who found that an increse in motivation has

11
a significant effect on the two approaches to level of aspiration. In
the "prediction“ group strong motivation led to a lowering of the dis-
crepancy score (more realism) while in the "hope" group, to a higher
discrepancy score (less realism).

Preston and Bayton (60) and MacIntosh (52) found that the ”least
level of aspiration“ (the lowest score one might attain) was lowered by
failure and even by reaching the level of performance (60) attained by
a superior reference group. The other two levels were not significantly
different, though both investigators did find that all of their subjects
tended to place the actual and maximum estimates closer to each other.

Holt (36) found results similar to Irwin and Mintzers. Controlling
success and failure and asking for three levels of aspiration, he noted
that in neutral and variable conditions subjects tended to give maximum
estimates a little more than four points better than their preceding
performance score. In success, however, this was markedly decreased to
only 2/3 of a point whereas in failure it was dramatically increased
to more than 18 points. "The least levels”, in keeping with the study
of Preston and Dayton and that of MacIntosh, were raised in success and
lowered in failure.

In summary, the level of aspiration seems to vary with success
and failure, either directly experienced in the present, or anticipated
as the result of broader life experiences. Thus, the negro subjects in
the study of Preston and.Bayton (60) lowered their levels of aspiration
even when informed that their performance equalled that of a white group.
As Gould (26) pointed out, the statement of the level of aspiration is a

complex phenomenon in which wishes and fears play a predominant role.

12

Reality-irreality also play a role in the type of task chosen. Both
Frank (19) and Gould (26) have found less consistency in'behavior in
tasks which are playlike in character. Bolt (36) (vide infra) interprets

these findings in terms of more or less ego involvement.
Forces Operating in Goal Setting

General Forces. The dynamics of goal setting have been analyzed by

 

several writers (1}, l8, “2, 51). (According to Frank.(18) "the level
of aspiration represents the final investigation of complex and con-
stantly shifting personal and situational factors” (p. 219). Different
investigators, notably Frank (17) and McGehee (53) have shown that the
goal the individual sets himself is not simply an intellectual estimate
of his performance but that conative factors play a part in determining
this estimate. Thus, if a subject is instructed to predict his score
in dart throwing (state his level of aspiration) while another subject
.is instructed to estimate the score which the first subject will ob-
tain, we find that "the estimator is more cautious in interpreting the
relation between the score and bids in any series of trials than is the
bidder.“ (52, p. 10).

Frank (16) states that the verbal statement of a goal is the re-
sultant of a conflict or choice between realistic judgement, the desire
for self protection, and the desire for gratification of one's self
esteem. The judgement factor would tend to approximate the goal to the
actual performance,the other two factors work to distort this judgement.
Of these two factors self protection tends to lower the goal so as to

increase the opportunities for exceeding it at the subsequent trial.

13
I The other factor however, works in the opposite direction; ittends to
push the goal up and is bound up with the individual's opinion of him-
self.

The preponderance of either of these two factors over judgement,
results in a deviation of the level of aspiration from that of perform-
ance, in either a positive or a negative direction. The extent to
which any of the three tendencies mentioned above will operate, will
largely depend upon the degree of the involvement of the ego-level

(Heppe's term).
Egg Involvements egg. Levels 9_f_ Aspiration

Both Hoppe (39) and Frank (11+) use this term to indicate the set-
ting of ideal goals and as self regard. Holt (37), in a brilliant re-
view of ego involvement, feels that this is rather confusing and states
that perhaps self regard is probably the most valid definition of
I'Ichniveau" (ego-level). ,

Though the ego level has been talked of since the beginning of
level of aspiration studies, little actual work has been done with it.
Sears (6“) states that "social values incorporated in the ego become
reference points for self judgements of success and failure, that is,
that a child cannot succeed or fail in an activity which.has for him
no ego value.“ (p. 526). In her study she found a few subjects who
were less involved in school work than were others. Three out of four
of these children gave inconsistent response patterns ihidh Sears felt

may be due to a lack of ego involvement.

1h

Heathers (32), though not primarily concerned with ego involvement,
did find that her highly motivated group (high school students who were
applying for a college scholarship) gave more consistent (more highly
intercorrelated) responses than did a group of college students paid
by the hour to'participate in an experiment.

Irwin and Mintzer (#2) may be considered.the first experimenters
to have controlled ego involvement. Their control group was told that
the experiment had not as yet begun and that they were in a practice
session; while the experimental group was instructed to do their best
in the dart throwing and were given a further incentive by being told
that their performance scores would be posted on a bulletin board. As
noted above the ego involvement group showed consistent results on two
levels of aspiration (the hope group giving higher discrepancy scores,
the prediction group lowering them).

Bavton (’4) operationally defined ego involvement by asking his
subjects which of two tasks (arithmetic and cancellation) they would
prefer to excel in. The task so chosen was defined as the ego involved
task. He found that a high expectation of future success was associated
with good subsequent performnce only in the ego involved task.

Bolt (37), in an intensive experiment with 10 subjects in which
both ego involvanent and success and failure were controlled, found that
level of aspiration behavior could be more successfully correlated with
personality traits when it resulted from relatively more ego involved
tasks. He states ”When more ego involved .... the aspiration loses its
close integration with other responses to the task at hand. It becomes
an expression of something deeper, more personal, less determined by

outer and more by inner forces.“ (p. 31h).

15
Though both Frank (11‘) and Hoppe (39) considered “ego-level“ as
both a combination of ego ideal and self regard, they tended to emphasize
the latter interpretation. Sears (bu) also tended to emphasize the ego
involvement in the latter terms - as the extent to which an individual
considers his performance as an expression of his own worth. Holt (38)
in an experiment involving predictions of future grades on examinations,
found results contrary to those of Beyton's (1+). He states that the
level of aspiration may be either goal striving or esteem defense; if
it is the former then one might deduce that a higher level of aspiration
should, if ability is held constant, be positively correlated with sub-
sequent achievement. If it is the latter, no relation should obtain.
He found that no correlation between aspiration and achievement existed
when past level of achievement was partialled out. He concludes (a)
When ego involvement is minimal, levels of aspiration have little
mgtihtional significance, being rimarily rational Judgements.
ch. Frank (17) and McGehee (53 (b) When ego involvement is
present but at low intensities, levels of aspiration have some do-
fensive meaning, but reflect to some extent the intensity of moti-
vation. (c) When ego involvement surpasses a certain limit, defen-
sive considerations become paramount and the level of aspiration
becomes more complexly determined. (pp. 1‘13-H1N).
The strength of the ego involvement seems to be a function of (a)
the task itself (b) the environmental factor or reference scales against
which the individual evaluates his performance (c) the personality make

up of the individual.

Characteristics 9; The Task

 

Festinger (l3), Escalona (10) and Levin gt gl_. (51) have pointed
to the fact that the level of difficulty of the task plays an important

part in goal setting. They state (51):

16
The greater the degree of difficulty .... the higher the valence of
success, within the boundary zone of ability. The absolute value
of the negative valence of failure usually changes in the opposite
direction. On levels of extreme difficulty the negative valence of
failure would be negligible. (p. 361).

Thus it would seem that a person would alWayS try for the most dif-
ficult task or consistently keep as high a level of aspiration as pos-
sible.

However, the question of probability enters into the picture. That

’ is, the chances of success or failure are evaluated in terms of certain

scales of reference, and the estimate of performance is formed accordingly.
External Reference Scales

Reference scales in terms of mthical or actual competitors have
been experimentally produced. In these experiments the scores obtained
by certain groups are given to the subject at some time during the task
(usually at first) and the level of aspiration is observed.

Anderson and. Brandt (3) in. an experiment with fifth grade children,
in which previous performance on given tasks was publicly posted, found
that there was a convergence of the level of aspiration toward the group
mean.

Those in the lowest ranks of achievement consistently set goals

considerably above their past achievements. Those in the upper

ranks of achievement set goals considerably below their preceding

achievement .... (Thus) .... the level of aspiration in all ....

achievement groups tended consistently toward mediocrity. (p. 232).
The same result obtained in the experiments of Hilgard, Sait and Hagaret
(33) with college students.

Gilinsh (25) informed college students about mythical performance

scores on a vocabulary test obtained by individuals with certain I.Q.'s.

17
She then had them state their I.Q. and the score they hoped to achieve
and compared the difference between the subjects own estimation of his
I.Q. and that of the comparison group, and its effect on the expected
performance on the test. She found as had Chapman and Vblkman(7) that
if the subject Judged.his ability to be superior to that of the comparb
ison group that the level of aspiration went up, if on the other hand
the comparison group was perceived as a superior the opposite result
was obtained.

Preston and Bayton (60) and Maclntosh (52) corroborated this with
white and negro students. Preston and Dayton working with groups of
negro students gave one group fictitious scores which white students
supposedly reached, and told the other group that negro students had
reached a certain score. They found that the group which was compared
to the whites lowered.their ”least estimates" in situations in which
they were informed they were doing as well as whites, significantly
more than the other negro groups when they were told they were doing
as well as negroes. They concluded that the need to avoid failure is
strongest when the students are compared to a socially superior group.
MacIntosh (52) reversed the situation by having a group of whites
”compete" with a mythical negro group. He found that the "highest“
and.'aetual levels“ (the most one might achieve and.what one Eggllz,
expected to do in the task) were raised but that the least level remain—
ed the same.

The expectancy of success and failure is thus well illustrated in
teams of external reference groups. If the group is seen as superior,
failure is expected in direct competition; therefore, the expected

achievement goal is lowered (7.25) or the expectancy of failure is

18

accepted via lowering of the least estimate of possible performance (60).
But even this behavior is not simply determined. We note (3, 33, 3h)
that if a person is too far behind the crowd.he tries to catch up, and
if too far ahead he slows down a little. As Holt so nicely puts it

(38) "a person asked for a level of aspiration is under social demands
of a conflicting nature: his statement should be realistic, expressive
of enough modesty, and yet show hhm to be properly ambitious." (p. N12).
The dominance of any one of these attributes seems to be a function of
that situation: realistic when one is told ahead of time that literary
critics have achieved a certain score on a test of "literary ability,‘
(7) modest when one sees ones-self conspicuously ahead of one's peers,
(3) ambitious and perhaps defensive when one is far behind (3) or when

one competes with social inferiors, (7).

Internalized Reference Scales

Other reference scales are internalized. Sears' (6“) very illumin.
ating study of successful and unsuccessful school children points to
significant differences in their level of aspiration behavior. Her
studies show that children who are academically successful have a rela-
tively low discrepancy score (i.e. their level of aspiration approaches
their level of performance) and show a high degree of responsiveness in

their level of aspiration to changes in the level of performance. Con—

versely, subjects whose achievements are below the average of their
group have a high positive discrepancy score (i.e. level of aspiration
is considerably above their past performance) and are less responsive

in goal setting behavior to changes in their performance level. Moreover,

19
their standard deviations as a group are larger, their goal setting be—
havior being less uniform than that of the success group. Gbuld (26)
obtained high discrepancy scores from individuals who came from a low
socio-economic background, while those who came from.uppernmiddle class
homes gave low discrepancy scores. Previous life experiences related
to success and failure are thus extremely important in goal setting and
seem.to act in much the same way as immediately experienced success and

failure.
Attempts Tngse Goal_Setting Behavior Qiagnosticallz

Other studies have tried to link level of aspiration behavior to

personality traits.

nggl_9£.lspiration and traits. The first systematic approach along this
line was pioneered by Frank (19) who attempted to correlate deviation
score and rigidity with personality traits obtained via a paper and pen-
cil test, He found that a high positive level of aspiration yielded a
correlation greater than.+0.20 with narcissism, aggression, emotionality,
creativity, projectivity, intensity, ‘Bigidity correlated better than
+0.20 with counteraction, impulsion, and endurance and -O.20 or better
with succorance and harm avoidance.

Previously Hoppe (39) had spoken of some relation of level of aspi-
ration behavior to ambition, prudence or daring, self confidence or
feeling of inferiority and courage to face reality. Jucknat (n5) sugp
gested a tendency to fear failure, ambition and prudence as possible
variables. However, both of these were speculative attempts to account

for the subject's bchavior in the level of aspiration situation.

20 .

Gardner (2“) tried to correlate the level of aspiration behavior with
ratings on subjective achievement level, dissatisfaction with status,
general sense of security, importance attached to intellectual achieve-
ment, tendency to face failure frankly, realism, fear of failure and
motivation. Though some trends appeared in his data he concluded that
no consistent relationship existed between level of aspiration and any
one of the variables dealt with.

In other words the mistake of Hoppe, Frank and.Jucknat was not in

suggesting that such factors were operative but in claiming that a
particular sort of aspiration level behavior could be directly and
unequivocally traced to the action of any one particular force v...
(the) level of aspiration behavior (rather) is due to the particular
organization or constellation of these factors in the given indi—
vidual. (p. 203).

Gould and Kaplan (28) in another attempt to correlate personality
factors conclude ”all that can be said at present is that what is elic-
ited in the 'level of aspiration' situation is not the same thing as
revealed by the questionnaires.I (p. 38).

More recently Klugman (MS) using the McFarland-Seits Psychosomatic
Inventory as a measure of emotional stability found trends which he felt
were ”high enough to indicate that there seems to be some .... relationp
chip between size of discrepancy and emotional stability as measured by
the Inventory." (p. 11%).

Lewin QEDEL: (51) state in relation to these correlational studies:
the evidence is far too slim.to provide a solid basis for future
thinking in this area. The variables so far investigated are prob-
ably too broad.and generalized to be usefully isolated as correla-
tives or determinants of specific level of aspiration scores such
as the goal discrepancy. (p. 351).

Holt (3b, 37), however, did find some significant correlations, in

ego involved situations, between level of aspiration behavior and

O 5

h

 

 

 

21
personality traits, which were obtained from a questionnaire ”Common
Forms of Behavior,‘I but only with ten subjects.

Rotter (63) has indicated that certain stable modes of reaction to
the level of aspiration situations do obtain. His analysis of extreme
scores suggested the presence of several patterns of response and also
suggested that the interpretation of the meaning of any score depended
upon the relation to other scores. Using deviation score, frequency
of shifts, unusual responses (lowering in success, raising in failure)
forgetting to state the bid, as well as an analysis of incidental be-
havior and comments he has arrived at nine definite patterns. In his
experiment he used fifty lbrcester State Hospital Employees, forty-five
College Students (both normal controls), twenty-one crippled College
Students, eighteen Prisoners (both experimental groups.) His patterns
which he states are not discrete include in order of descending adjust-
ment level: (1) low positive discrepancy pattern, (2) low negative or
very slight positive deviation, (3) medium'high deviation, (H) achieve-
ment followers, (5) step pattern, (6) high positive deviation, (7) high
negative deviation, (8) rigidpattern, and (9) confused or breakdown
pattern.

He concluded that the full value of this technique for clinical
and some experimental purposes lies in the consideration of the whole
pattern of response rather than of the deviation score or any other single
score. Rotter feels that his results indicate that the test roughly dem-
onstrates the degree of feelings of inferiority or inadequacy and the
nature of the defense or compensation with which the subject attempts to

meet his feelings of inadequacy. He considers the test to be clinically

22
diagnostic to a high degree at the extremes. Diagnostic accuracy within
the group of less extreme responses appears reasonably high but further
work is necessary to determine this more accurately.

In general it may be said that attempts to correlate personality
traits arrived at by paper and pencil tests with results obtained in a
level of aspiration setting have not proven their worth. Perhaps it is
as Gardner (2%) states that level of aspiration variables are too broad
and generalized to be correlated with personality inventories. But the
fact that these inventories are far from.the best indicators of personp

ality characteristics, should also be considered.
Studies With Clinical'Groups

a.more feasible approach may be via clinical groups where somepre-

dictions in terms of known or hypothesised dynamics may be made.

Studies with neurotics. Eysenck and Himmelweit (11) and Himmelweit (35)

 

have conducts! experiments with two groups of neurotics, and have com-
pared these to normals. In the first experiment (11) two groups of neup
rotics; hysterics and affective syndromes (differentiated previously by
a factorial study) were compared.

Following Jung's typology they hypothesized that the introverts'
(affectives) level of aspiration should be less affected by success and
failure than that of the extroverts (hysterics). Using a modified pur-
suit rotor they found that (a) affectives show greater interpersonal
variability than.hysterics, (b) hysterics show greater intrapersonal
variability, (c) affectivestend to keep the level of aspiration at the

same level after both success and failure whereas hysterics did not.

23
They conclude that the affectives (anxiety, depression, obsessional,
irratable or apathetic cases) were more rigid and subjective, neglecting
external reality whereas hysterics placed their emphasis on objective
and factual relations. In a comparison with normals Himmelweit (3h)
found that the neurotic scores spread over a wider range than did the
normals; there was a bi-modality to the neurotic groups' scores (cf.
above). Both groups however overestimated their level of aspiration,
also both underestimated their past achievement, the difference between
the groups being quantitative rather than qualitative. In general ,
normals were more realistic than neurotics, placing their level of as-
piration somewhat in keeping with their estimation of their past per—

formance.

Studies with psychotic 532223, Escalona (10) was the first to attempt
an investigation with psychotics. She administered h series of paper
and pencil mazes, controlling for success and failure by Jucknat's (HM)
method, to 16 manics,,2h depressives and a control group of 33 college
students.

Although she states that the number of cases analyzed were inade-
quate for statistical purposes, certain trends did appear. The de-
pressives kept on the same level of aspiration for all tasks, success
and failure alike. She felt that they were less influenced by the char-
acter of the situation but by inner forces which changed little during
the experhment. The manics though aspiring high at first lowered their
level of aspiration after the first encounter with failure. She stated
that the manics approached the task.superficially, and in accordance

with their moods and increased self esteem, they tended to shoot high

21$

simply because they never considered failure Possible. However, only
one encounter with failure was enough to lower their level of aspiration
and to keep it there.

A more recent study by Miller (57) with four clinical groups (con-
version hysterics, character disorders, paranoid schizophrenics and
neurasthenics) merely adds to the already existing confusion. His find—-
ings with the neurotic groups are almost contradictory to those of
Eysenck and Himmelweit (11). Using a three dimensional form board, the
Carl Hollow Square Scale, Miller found that the conversion hysterics
were “the most defeated group. The mean goal discrepancy is lowest; it
has the fewest S’s with positive median goal discrepancies and the few-
est with a mjority of positive individual goal discrepancies, (and
further) the conversion hysterics consistently predict that their scores
will go down.'' (57, p. 386).

This is in contrast to Eysenck and Himmelweit's statement that
hysterics as ”extroverts" are sensitive to the environment and will ad-
just their levels of aspiration to the success and failure of the pre-
ceding trial. Miller's‘neurasthenic group though not exactly equivalent
to the so-called affectives shows many of the latter's traits (anxiety,
depression, apathy), and so may be roughly compared. Again we note
discrepancies. Whereas liysenck's effectives were relatively rigid and
unresponsive to external reality (i.e. to success and failure) keeping
their levels of aspiration at virtually thesame level throughout the
whole of the experiment, Miller's neurasthenics ”find it difficult to
predict anything but improvement." (p. 385).

The writer is unable to account for these diametrically opposite

results with only the mblished data available to him. However some

35
differences in method (e.g. induced degree of motivation, control of
success and failure) as well as selection of the population may account
for the results. In a way this highlights the great difficulty of exper.
imentation with goal setting behavior. The amazing thing is not that
the literature sometimes yields contradictory results, but that in the
assessment of so complex a phenomenon we have had the overall consistency
obtained up to the present time.

With character disorders Miller again found rather consistent re-
actions. He stated that "the character disorders .... appear to be the
most realistic group” (57, p. 386) in terms of level of aspiration. How-
ever, they "reveal considerable conflict and anxiety in the face of pos-
sible failure e... they seem to be so fearful of failure that they can-
not risk predicting improvement even when such a prediction is warranted."
(57. p- 386).

His normals like Eysenck and Himmelweit's are the most responsive
of the experimental groups. They exhibit a slightly positive discrep-
ancy score and "compared to the psychiatric S's (they) seem less over-
whelmed by the possibility of failure.‘ (p. 386).

Studies using schizophrenics as subjects are almost as rare as
those using other clinically diagnosed groups. Hausman (31) in an
early study with a diverse group of psychiatric patients found that
paranoids showed a rigidity and a stubbornness in clinging to a rather
high level of aspiration despite any change in performnce. He does not
report any other significant observations, with paranoids and he apparent-
ly did not investigate any other schizophrenic groups.

Recently Bowman (5) and Lyle ("-9) have attempted more quantitative

studies with groups of schizophrenics. Both attempted somewhat unsuccessfully

 

cee'

26
to correlate goal setting behavior with the Rosenzweig P. F. test. Howb
ever, both did find that in general schizophrenic patients tended to
show a significantly higher positive goal discrepancy score than did
the normal subjects, and further that schizophrenic patients showed some
trends toward rigidity (i.e. they were less responsive to previous ex-
periences of success and failure than were normals.)

Bowman used a group of thirty schizophrenic patients in her experi-
ment. They were randomly selected from a group of male patients who
were under forty years of age and who had been.hospitalized for less
than five years. The sample consisted of 10 patients diagnosed.hebe-
phrenic, 1 simple, 9 paranoid, 3 catatonic and 7 mixed schisophrenics.

A control group of 30 males picked randomly from the hospital staff was
alsO'used.

The task consisted of rolling wooden balls so that they would sink
into cups on a board twenty feet away. No attempt was made to ego-in-
volve the subjects, nor was there any control of success and failure.
She found (a) that the discrepancy between the level of aspiration and
the level of past performance was significantly higher for her schizo-
phrenic patients than for her "normals“, (b) that while no difference
existed between performance scores, the schizophrenic group's level of
aspiration score was Significantly higher than that of the “normals”,
(c) the aspiration and performance scores of the control group were
somewhat highly correlated (mean r.u65) whereas no such correlation
existed for the schizophrenic group (mean r.030), (d) that the schizo-
phrenic group showed some tendency toward more rigidity, i.e. invariance
of iesponse, than did the controls, though this did not reach statistical

significance.

27

Using the Rosenzweig Picture Frustration test she found that no
significant difference existed between the schizophrenic and control
groups on any of its nine factors. She did find a correlation of 0.322
between the group conformity score on the Rosenzweig and the average
D score for normals while the schizophrenic group's D score correlated
at a non-significant level (0.115).

She also found that the so-called super-ego score (extra and intro
punitiveness scores combined) correlated significantly with D for the
schizophrenic groups but not so for the controls. However, she made
no attempt to account for these results, rather dismissing the Rosenz-
weig Test as a probably invalid instrument.

Kyle (M9) performed essentially the same eXperiment with better
controls. His population consisted of 6h male schizophrenics (16 of
each type) and 6M normals matched for age, education and socio-economic
background.

The subjects were presented the Rotter Jensen Formpboard ( a board
with a series of depressions over which the subject shoots a ball bear-
ing by means of a short cue), as a task of motor coordination. Failure
was controlled by means of surreptitiously raising the angle of the
board in such a manner that high scores were made impossible; success
however, remained uncontrolled.

Kyle's findings agree essentially with Bowman's. He found (a)
that there were no significant differences between the beginning levels
of aspiration for normals and schizophrenics, (b) the schizophrenic
group as a whole showed a.higher deviation score than did the normals,

(c) the schizophrenic group showed a higher level of aspiration than did.

28

the normals, (d) the schizophrenic group was more rigid than the control
group but not at a statistically significant level, (a) the control
group generally raised their level of aspiration after success and lows
ered it after failure while the schizophrenic groups remained at the
same level or raised after failure as well as after success, (f) para-
noids and hebephrenics were the most rigid of the schizophrenic groups

‘ staying on the same level despite any change of fortune, (g) paranoids
in general had a significantly higher level of aspiration than did the
other schizophrenic groups.

His results with the Rosenzweig P. F. test are also similar to those
of Bowman in their lack of clarity. He found only three intercorrela-
tions to be significant for the schizophrenic group (-0.57 in the para—
noid group between average D and general conformity,-.lbl in catatonics
between average D and general conformity, and -O.50 in simple schizo-
phrenics between average D and need persistence).

no significance was found in the normal groups between the Rosenz-
weig and goal setting behavior.

Essentially then, these two studies are in agreement. Kyle's con-
clusions are very similar to Bowman's: (a) the high discrepancy score
is interpreted as a rejection of reality, and also as related to anxiety,
insecurity and a general lack.of self confidence, (b) the rigidity noted
is interpreted in terms of the rigidity often seen in schizophrenic be-
havior.

Miller's (57) results with a group of paranoid schizophrenics more
or less corroborates the work.of Kyle and Bowman. While the paranoids

as a group show only a low positive mean discrepancy score and thus

29
resemble the normal controls, "the paranoid schizophrenics are compar-
able to the neurasthenics in their devotion to high attainment ....
when the reality situation requires prediction of the same or a lower
score, the schizophrenics still feel compelled to estimate some im-
provement.‘ (p. 385).

Thus it seems that somewhat consistent results have been obtained
in the studies of goal setting with schizophrenics. All investigators
but Miller found rather high positive goal discrepancies in their
schizophrenic groups. All have found some sort of lack of responsive-
ness to the demands of reality. But as yet no one has been able to at-
tribute these behaviors specifically to failure. Of the four experiments
cited only Miller's was designed to make it possible to abstract post
failure behavior from pre—failure behavior. Despite this, Miller does
not specifically make the needed analysis.

None of the four investigators have studied the effects of differ-
ent degrees of failure with schizophrenic patients.

None of the four investigators have studied the effects of failure
on perfbrmance.

For these reasons a study which can control success and failure
in an objective way and which can further control degree of failure
seems to be called for. Only in this does it seem possible to begin
to establish or disprove whether the behavior noted in goal setting

situations is explicable in terms of reaction to failure.

M
The basic problem posed by this discussion may be stated as fol-

lows: lhat effect does failure have on the goal setting bahavior of a

30
group of schizophrenics? It has been noted.that it is difficult if not
impossible to compare directly life goals of individuals or groups, but
that one might obtain information bearing on the question posed via a
level of aspiration experiment.

A brief history of the method, of its special characteristics and
of the attempts to use it to further explore personality, indicate that
more recently,better controlled studies of goal setting behavior have
been carried out. The method, at first quite qualitative, has become
more quantifiable. As a result of this, studies with various groups are
more feasible and more meaningful. Despite this, comparatively little
work has been done with schizophrenic patients.

The studies which have been published consistently indicate a higher
level of aspiration and a greater degree of rigidity among schizophrenic
patients, but none have as yet attempted to link:these directly to fail-
‘ure, nor have any attempted to control the degree of failure.,

It is to this task (the relation of goal setting behavior to failure)

that this study addresses itself.

SCHIZOPHRENIA FAILURE AND FHUSTRATION
Theories _o__i_'_ Schizophrenia

The problem of achieving a thorough understanding of the schizo-
phrenic processes still remains unsolved. The thinking on the nature
of schizophrenia has ranged from.the completely organic approach pos-
tulated by Kraeplin to the more environmental approach advocated by
Adolf Meyer and.more recently by Cameron. The Freudians seem to hold
an intermediate position between the two. Abraham (1) feels that there
is a congenital psychosexual constitution for Dementia Praecox, which
predisposes the individual toward withdrawal in the face of conflict.
He emphasizes that unlike neurosis where the conflict is internal, the
battle of the psychotic person is between the ego and the outer world.

Freud (21), writing somewhat later than Abraham, plays down the
l'psychosexual constitution” though he does not abandon it. He states,

There always remains as a common feature in the etiology both of
the psychoneuroses and the psychoses the factor of frustration.-
the lack of fulfillment of one of those eternal uncontrollable
childhood wishes that are so deeply rooted in our composition,
phylogentically foreordained as it is .... Now the pathogenic ef-
fect depends upon whether in,the tension of such a conflict the

ego remains true to its allegiance to the outer world and endeavors
to subjugate the id or whether it allows itself to be overwhelmed
by the id and thus torn away from reality. (p. 253).

Fenichel (12) who summarizes the major analytic work up to recent
times stays quite close to the above formulations. He feels that frue—
tration is an important aspect, but that unlike the neuroses where

frustration alone is operative, in schizophrenia both constitution and

precipitating experiences play complementary roles.

32

Cameron (6) feels that schizophrenia is a disorganization of the
biosocial reaction system which characterizes personality. Lacking
ability in role taking as a function of early difficult childhood ex-
periences, the schizophrenic, faced with a series of serious life prob-
lems, cannot adequately deal with.them, and slowly retreats beyond the
reach.of the world and its frustrations.

Recently Jenkins (M3) has postulated the hypothesis "that schizo-
phrenia is a progressive maladaptation resulting from.frustration be-
yond the tolerance of the patient." (p. 256).

Using Maiers (55) theoretical framework as a basis, Jenkins feels
that continuous frustration beginning in childhood "inevitably leads
to further frustration, the process [being7'typically progressive.“ {h},
p. 256). He cites numerous references which point to the fact that the
schizophrenic patient has as a child been either rejected or overpro-
tected by the parents who are described as "overpowering parent(s) with
whom it would be more than.usually difficult for a child to establish
his individuality, his self, without doing so through negativism or
schizophrenic withdrawal.“ (1+3, p. 2’46).

Cameron (6) states essentially the same thesis when commenting on
the growth and development of schizophrenic patients, namely that the
parents are either rejecting or overprotecting. This behavior leads to
the development of self attitudes which lead the individual to become
unusually "reaction sensitive” to signs that he may be guilty, inferior
or unworthy in the eyes of others. That is to say, that as a result of
these acquired self attitudes they may be more ready to react to stimuli

in the environment in a manner consistent with these attitudes. Thus,

33
derogatory or unfriendly comments are more readily perceived by persons
who are "reaction sensitive" to signs of inferiority and guilt. As
time goes on a "progressive reaction sensitization” occurs, i.e. a
"process in which a person, once he has become reaction-sensitive in
some specific direction, continues to develop further readiness-to-
react in the same direction on the basis of successive acquired reactions."
(6, p. 69). Thus, in time even neutral stimuli may be interpreted in
terms of this basic frame of self reference, leading to delusions, hal-
lucinations and other clinically observed abberations.

Fromm-Reichmann (22) states:

traumatic experiences in this early period of life will damage a
personality more seriously than those occurring in later childhood
such as are found in the history of psychoneurotics. The infant's
mind is more vulnerable the younger and less used it has been; fun-
then trauma is a blow to the infant's egocentricity. In addition
early traumatic experience shortens the only period in life in
which an individual ordinarily enjoys the most security thus endan-
gering the ability to store up, as it were, a reasonable supply of
assurance and self reliance for the individual's later struggle
through life. Thus is such a child sensitized considerably more
towards the frustrations of Later life than by later traumatic
experience. Hence many experiences in later life which would mean
little to a healthy person and not much to a psychoneurotic mean

a great deal of pain and suffereing to the schizophrenic. His re-
sistance against frustration is easily exhausted. (p. ul3).

§EEEE£23 These authors seem to view the schizophrenic in terms of early
childhood frustrations and failures which leave the individual "reac-
tion sensitive“ to many more frustrations. The sensitivity leads to an
increase in frustration, the frustration leads to maladaptive behavior
which.may be seen as defensive. This behavior by definition does not

gain.the effects which are desired fromtthe world and leads to greater

frustration and so on, cyclically.

3h

The Effects q£_Frustration qn_Goal_Directed Behavior

One of the most potent areas of frustration lies in the field of
interpersonal relations, particularly in setting and achieving goals.
As Alexander (2) has pointed out, our children are filled with stories
of great accomplishment and ambition, with the feeling that the world
belongs to anyone who will get out and work, but the reality of this
frontier philosophy cannot be easily realized. The results can often
be disasterous to the adolescent and young adult. In this connection
Adolf Meyer [footed in Jenkins (1432] states:

The greatest difficulty in life, the greatest source of disharmony,
apart from.the influence of heredity, infectious disease and poor
feeding and poor chances of growth, is the discrepancy between im-
pulse yearning and ambition, on the one hand and the actual opporb
tunities and the actual efficiency of performance on the other. We
know people who try continually to put square pegs into round.holes.
They are unable or unwilling to learn to know and to accept their
own nature and the world as it is, and to shape their aims accord-
ing to their assets.

In a large percentage of cases in which persons come to grief in
their mental and moral health, the trouble is of just that kind.
Failing with what is frequently impossible and undesirable anyhow,
these persons develop emotional attitudes and.habits and tendencies
to fumble or to brood or to puzzle or to be apprehensive until what
students of the functional diseases of the head call 'a break of
compensation' occurs, a break of nature's system of maintaining the
balance with a more or less sudden slump and implication of collat-
eral functions.

It is known that children who have experienced continuous failure
before, act unrealistically in an experimental goal setting situation
(6”), either setting the goal far too high, striving as it were only to
create an impression of their ambition, or even perhaps living in a
dream world in which the statement of the goal is magically equivalent
to its attainment; or setting the goal ridiculously low and gaining sat-

isfaction from the fact that they are not as bad as all that. Preston

35
and Bayton's (60) negro subjects show similar sensitization to defeat
presumably as a function of the conviction of their inferiority to whites.

Studies with those who apparently were not sensitized to failure
(inferiority) seem to indicate the same unrealistic defensiveness when
the stigma of failure is inevitable. Sears (6“) found that in a severe
failure situation:

the general hostility level was less for the failure than for the

success subjects. The frequency of daydreaming and autistic thinkh

ing was sharply increased, and the social responsiveness (was) reduced

.... All this reduces the possibility of his having new experiences

or of initiating new ways of behaving. He avoids the environment.
Second in direct relation to this, failure leads to a dogged but

ineffectual continuation of the task at which failure occurred.

What interaction with the environment there is,is in the direction

of the old activity. But the old activity is half avoided, and the

card sorting is unfinished in order to avoid the danger of failing;
this effectively precludes success, and therefore, the person fails

anyway ....
Finally, the process of deoontextualization that failure subjects

exhibit serves in still another way to reduce their adjustive effect-
iveness. This process splits off the activity from its social frame
of reference, reduces its contact with reality, and hence decreases
the opportunities for the person to check upon the task's hmportance
by reference to reality. (pp. 256-257).

Kendig (#7) testing a group ofcollege students on the production
of words beginning with the letters ”c" and ”s” found that under experi—
mentally induced failure, there was more perseverative activity. Lantz
(50) found significant differences between a previously equated success
and failure group on the Binet Form.M. She concluded that the experience
of failure in an unrealted activity seems to have a great effect on men-
tal processes.

Katz (#6) in.a recent failure experiment found that subjects who
failed drew human faces with a great deal less expression. He explains
the drawings and the incidental failure behavior in terms of “(1) cover-

ing up or self concea lment (to) insulate the person against the power

36
field of the thwarter (2) apathy or depressed psychological functioning
(which) reduces emotional tension and lessens awareness of the implica-
tion of failure." (p. 3H8). He states that in failure,”Emotional ten-
sion is reduced as failure aroused attitudes and feelings lose their
sharpness and intensity. And as cognitive processes become dulled.the
painful significance of failure is lost. Secondly conmunication with
the external environment is reduced, i.e. the person is both indiffer-
ent and inactive." (p. 3n7).

Marquart (56) duplicating Maier's rat experiment with human sub-
jects found that some of her subjects became fixated while other did
not. She explained this bimodality in terms of threshold, i.e. that
fixation occurred only in those cases where the impact of punishment
exceeded the frustration threshold.

This threshold concept has been elaborated by Eglash (9) who
states:

according to this principle there is some thing we can term.as a
limit or capacity which organisms possess for tolerating or endur-
ing increments of excitation. As accumulative increments mount,
quantitative changes pile up until the limit of the animal's ca-
pacity to contain them is reached. lhen threshold level is reached,
something different from what has been happening before then occurs.
There is a quantitative change. (p. 8%).

French (20) in a theoretical paper lists several prevalent reac-
tions to frustration: (a) concentration upon overcoming the obstacle;
here the goal remains unchanged, new energy is expended, renewed effort
is tried. "at least temporarily the original goal is forgotten. The
energy of the original wish is concentrated.upon the task.of overcoming
the obstacle. Not the original goal but the obstacle becomes the center
of interest.“ (p. 65) (b) Deflection to avoid the obstacle; this may

include getting around it or'modifying the goal.

37

These French labels successful attempts to deal with non attain—
ment of a goal. The unsuccessful attempts do not differ in kind from
the above but only in intensity. Thus concentration on the obstacle
may become fixation on the obstacle and deflection, neurotic substitup
tion.

This difference in the effects of failure may be seen in HurlOCk's
(Ml) study. She found that reproof as well as praise acted as an inp
centive and both increased efficiency of performance when compared to
a “no comment" situation. However continued reproof over several days
led to a disintegration of efficiency. Steisel and Cohen (67) control-
ling degree of success and failure found that in a setting of success
even strong failure is motivating, as a function of "achievement expec-
tancy.” In other words, if the subject is accustomed to success,
whether it be in life experience or in an experimentally produced situp
ation, an occasional bump on.the head can spur one on, if failure has

been the mode it can only lead to non-adaptive defensive behavior.

Summagz, Many theorists include frustration as one of the possible
factors in schizophrenia. In an achievement oriented culture such as
ours failure to obtain one's goals can be, as Meyer [Eh Jenkins (M3)].
points out, a major cause of frustration. If as Cameron (6) and Jenkins
(#3) assume the schizophrenic, as a function of many many early frustra-
tions, is "reaction sensitive" to them, has a lowered threshold as it
were, it is possible to conclude that a little frustration will go a
long way with schizophrenic patients.

It has been noted, particularly in Sears' study, (65) that appar-

ently "normal“ individuals will become withdrawn, rigid, and perseverative

38
in the face of strong failure. However, if the failures are tempered
by success as in Steisel and Chen's study (67) they apparently can be
tolerated and may even prove to be motivating. Thus it seems to be a
matter of degree not of kind. This is not unlike the reaction of two
persons, one of whom is sunburned, to a slap on the back. (A slight tap
will not disturb the unbaked person, while it might result in consider-
able pain for the other. However, a rather lusty blow will even effect
the latter.

Thus with frustration, it seems possible that slight amounts will
elicit rather unrealistic behavior in schizophrenics, but not in normals.
If it can be assumed that failure to achieve a stated goal is frustrating
we can change the above proposition into a more testable form by substi-

tuting the word failure for the word frustration.

Hypotheses

The first major hypothesis of this study can therefore be stated
as follows:
Hypothesis I. After failure the schizophrenic subject will be less in-
fluenced by his past performance in goal setting than will the control
group.
Thus:
Hypothesis a. The schizophrenic subjects will tend to give a.higher
level of aspiration following failure than will the controls.
Hypothesis b. The schizophrenic subjects will tend to give higher levels
of aspiration following failure than in success.
Hypothesis c. The schizophrenic subjects will give higher levels of as-

piration following strong failure than they will following mild failure.

39
Hypothesis d. The schizophrenic subjects will tend to give a higher
positive discrepancy score (level of performance of the previous trial
minus level of aspiration for the next rial) after failure than will
the controls.
Hypothesis e. The schizophrenic subjects will give higher positive D
(discrepancy) scores following failure than they do in success situa-
tions.
Hypothesis f. The schizophrenic subjects will give higher positive D
scores following strong failure than they will following mild failure.

The second.major hypothesis may be stated as follows:

Hypothesis II. As a result of the postulated reaction sensitivity to
failure the quality of work of the schizophrenic subjects should be
more disrupted than that of the controls.
Therefore:
Hypothesis a. The schizophrenic subjects will show a greater drop in
efficiency (seconds taken per item) following failure than will the
controls.
Hypothesis b. The schizophrenic subjects will show a greater drop in
efficiency after failure than after success.
Hypothesis c. The schizophrenic subjects will show a greater drop in
efficiency following strong failure than they will following mild fail-
ure.
Further:
Hypothesis d. The schizophrenic subjects will show a greater number
of errors following failure than do the controls.
Hypothesis 6. The schizophrenic subjects will show a greater number

of errors following failure than after success.

ho
Hypothesis f. The schizophrenic subjects will show a greater number
of errors following strong failure than after’mild failure.
Hypothesis III. The schizophrenic subjects will attempt to withdraw
from.the failure situation to a greater extent than will the controls.
Hypothesis a. The schizophrenic subjects will take a longer time to
state their next level of aspiration after failure than will the con.
trols (this is referred to as longer reaction time).
Hypothesis b. The schizophrenic subjects will show a longer reaction
time after failure than after success.
Hypothesis c. The schizophrenic subjects will show a longer reaction
time after strong failure than after mild failure.
Hypothesis d. The schizophrenic subjects will attempt to deny a sub-
jective feeling of failure more often than will the controls.
Hypothetis e. The schizophrenic subjects who experience strong failure
will attempt to dany a subjective feeling of failure more often than

those who experience mild failure.

METHODOLOGY

Reguirements gf_the Experiment

 

As has been noted before, the results obtained in several studies
which involved level of aspiration behavior have been unclear or even
contradictory. Part of this lack of clarity must be attributed to the
inadequate use of experimental controls. There can be no doubt that
goal setting behavior even in a quasi-artificial experimental situation
is extremely complex. Since the effect of many factors on goal setting-
are as yet unexplored, they must somehow be held constant, or at least

minimized.

Requirements ig_the selection of the groups. Several factors seem to

 

 

be of great importance in the selection of the group: (a) Age - Though
no experimental evidence has as yet been produced showingua direct con-
nection between age and level of aspiration, the possibility that a
shift in levels of aspiration occurs with aging cannot be ignored. (b)
Education - This seemed to be of importance in this study since one of
the tasks required the solution of arithmetic problems. Even though
these problems were of a relatively simple nature, it seemed as if edu-
cational background might make a difference in terms of ability and con-
fidence with which the problems were approached. Since initial levels
of aspiration are often a function of past performance 16f. Kyle (M9)
and Bowman (5)]; any difference in ability and/or confidence might re-
flect itself on goal setting behavior. (c) Intelligence —- This was

also a possible factor to be considered in relation to the first task

he

(arithmetic). Again its effect on ability in the practice trials and
indirectly on goal setting had to be controlled. (d) Initial Perform.
ance Ability on the Tasks - As has been noted this seems related to
initial levels of aspiration. (e) Degree of Reality Contact -- All of
the schizophrenic subjects used.had to be in sufficiently good contact
(as determined by a brief interview) to follow the given instructions.

(f) Lack of Neurophychiatric Disability in the Control Group.

Requirements in_the selection gf_the task. The criteria for tasks to

 

be used in a goal setting behavior experiment have been listed by several
writers, among them Holt (3S), Rotter (62) and Gardner (23). They may
be summarized as follows:

a) The task should be able to pass as a measure of whatever ability
the examiner claims it measures.

b)~ The task should be able to be divided into units of performance
so trends can be easily seen, '

c) The nature of the task should be such that the primary emphasis
can be put on time rather than on achievement or errors, since the sub-
ject finds it much harder to judge time accurately,

d) Brief trials should be feasible so as not to tire the subject,

e) The task.should be interesting enough to guarantee sustained
attention.

Some controversy exists about the novelty of the task, Better (62)
and Gardner (23) both considering thiI of importance. However, Holt
states: "To have used tasks of the neutral or colorless character that
Rotter and Gardner demand would.have been to insure that they were re-

moved as far as possible from ego involvement." (38, p. 92).

u}
These then are the prerequisites which were considered by the writer
in selecting the groups and the tasks. Undoubtedly more uncontrolled

factors exist but only further research can bring them to light.

Thg_Groups

E22i2.2£.£§21£2.2£.th3 control group. As noted above the basic criterion
for inclusion in the "normal“ group was a negative psychopathological
history. Since the members of the control group were allpatients at a
Veteran's Administration General Medical and Surgical Hospital, it was
relatively easy to obtain case histories. The case history along with
an opinion obtained from the patient's physician constituted the basis

of choice. The following factors were considered: (a) no history of
hospitalization for a neuropsychiatric disorder, (b) no hospitalization
for a psychosomatic complaint, (c) no history of encephalopathy, (d)
negative history of accident proneness, (e) opinion of the attending

physician as to current absence of psychopathology.

§g§i§_g£_ghgi£g_gf_the schizophrenic group. The schizophrenic group
posed a few more problems in selection. Since the hypothesis concerned
the behavior of schizophrenics in general as opposed to specific sub-
categories, the writer faced the Problem of getting a representative
group. Three possibilities existed (a) Bowman's (5) method of obtain-
ing a random sample from the hospital regardless of subcategory, (b)
Kyle's (49) method of obtaining individuals representative of each
category, (c) obtaining a diagnostic sub—group which might show a cross
section of schizophrenic pathology. The last method was chosen, and a

group of mixed or unclassified schizophrenics was selected. There were

mi

several reasons for this: (a) the population of mixed schizophrenics
was one of the largest in the hospital with the possible exception of
the paranoids, (b) with the exception of paranoids this group had the
greatest number in good contact, (c) this group by definition (58, 67)
showed.a mixture of behavioral manifestations seen in all of the other
categories, (d) to an extent one had to worry less about “pure" cases
here than in any of the other groups, thus making criteria of choice
simpler.

The paranoid group was not used, as it seemed least representative
of the schizophrenics both in terms of age, intelligence, etc., [;f.
Cameron (6), Noyes (58), Hunt and Cofer (“017 as well as in terms of
goal setting behavior [Ef. Miller (57)]; ‘A representative sample of
each subcategory was rejected since: (a) Kyle (M9) and Bowman (5) in-
dicate that in.general there is a much greater difference between schizo-
phrenics and controls than among schizophrenics thus making it feasible
to test only one group, (b) Hebephrenics and catatonics in good contact
were not available in sufficiently large numbers, (c) Clearly defined
diagnoses of schizophrenics subgroups by means of objective or projec-
tive tests is not feasible or at least has not been conclusively demon—
strated.

A randomized selection was avoided since if the criteria of good
contact is used a preponderant amount of paranoids will accumulate. (wide
supra). The specific criteria utilized for the choice of the schizo-
phrenic group, other than the normative data were (a) diagnosis of schizo-
phrenia type unclassified after review of the case by the hospital staff

(psychiatrists, psychologists, social workers). (b) Agreement of the

1*5
results of the Rorschach with this diagnosis. [3%. Schaefer (p. 78171
(c) Lack of any definite projective tendencies of long duration, as
determined by the history. (d) Relatively good contact with reality

(as determined by a brief interview previous to testing).

(The normative data. For purposes of the experiment the schizophrenic

 

and control groups were both divided into two subgroups of 16 persons
each. An attempt was made to match all subgroups as to age, education;
intelligence, and performance ability. Intelligence was measured by
means of the C.A.S., a short form of the Wechsler consisting of the
Comprehension, Arithmetic and Similarities subtest (Ff. Rabin (60) for
reliability data on this short fomp7. Performance ability was measured
by the level of performance reached on the last of the five practice

trials with each task (vide infra). The results are presented in Table 1.

TABLE 1

A COMPARISON OF THE GROUPS WITH RESPECT TO SOME NORMATIVE FACTORS

 

 

 

 

 

 

Schizo. Schizo. Control Control

Group 1 Group 2 Group 1 Group 2
Age M 26.2 26.8 25.9 26.7
S.D. 5.29 6.03 5.84 5.28
I.Q. M 102.5 101.9 102.5 101.h
S.D. 5.20 5.ou than 5.72
Ed. M 10.2 10.2 10.1 10.1
S.D. A 1.59 1.51 1.54 1.68
Perf.

Task I 9.87 9.63 9.62 9.93
S. D. '2.20 2.36 2.38 2.36
Perf.

Task II 9.00 8.9% 9.06 9.43

S.D. 1.92 1.52 1.52 1.15

 

M6
As can be seen the groups are fairly close to each other in all
of the factors which we considered. This is no doubt due to the manner
of selection. The following normative criteria were employed in the,
group's selection: (a) Age range of 19-38 years. (b) I.Q. range of 90-

110, (c) Educational range of from 8-12 grades.
The Tasks

The Arithmetic Problems. Twenty-five sets of 30 simple two digit addi-
tion problems were used for the first tack. The problems were constructed
with the aid of a table of random numbers. To further aid in randomiz-
ing possible difficulty one problem was written on each of the twenty-
five pages which were then shuffled. The second problem was then written
on each page, and the pages were again shuffled. This continued until
there were thirty problems on each page.

The use of thirty problems was arbitrary. However, the following
considerations had to be taken into account. (a) There should be a suf-
ficient number of problems to offer some sort of a challenge to the
subject when told he was to complete as many as he could in a given time,
(b) the number of problems should be suffieciently large to allow for
the possibility of continuously increasing the level of aspiration over
all of the trials, (c) on the other hand, there should not be so many
problems that the subject would be discouraged from the start. »In gen-
eral the use of twenty-five trials of 30 arithmetic problems each meets
all of our stated criteria for the task. They were introduced as a test
of intelligence, and the writer feels that in most cases they were accepted

as such by the subjects.

n7
Thg_§gggg§_t§§ka‘ The one objection which can be leveled against the
use of arithmetic problems is that arithmetic is often an emotionally
loaded task, and that there might well be definite intragroup and/or
intergroup variability as a function of the subject's attitude toward
arithmetic. This might well effect our other measurements. In order
to control for this, a second task was also presented. This consisted
of a modification of the Johnson O'Connor Tweezer Dexterity Test (59).
This test consists of a board in which small holes have been drilled.
The object is to lift a metal pin from a tray by means of a tweezers
and insert it in the hole, starting with the lower right and.working
each row from left to right. The original test consists of one hundred
'holes (10 rows of 10). Since brief trials were considered essential
for our study, seven of the ten rows were covered by a piece of card-
board leaving thirty holes available. This task was presented as a

test of mechanical aptitude, and seemed to be accepted as such.

Time lhmits. It is obvious that if the subject were given unlimited
time in either of the tasks he could easily complete them. The time
limits had to be set in such a way as to allow the individual to do
several problems or put in several pins, thus encouraging him, but they
also had to be set in such a way as to prevent his completing all of
the problems in the first few trials, thus leaving nothing more to as-
pire to.

The determination of the optimal time for our groups had to be on
an empirical basis. At first the thirty second time limit used by
Steisel and Cohen (67) was tried with the arithmetic problems. The sub-

jects (five randomly selected schizophrenics and five matched attendants)

1.3

finished from 8 to 15 problems (M 12.5) on the first trial. This was
felt to be too many problems so a time limit of 20 seconds was used on
a second group of five schizophrenics and five attendants. The mean in
this case was about 8.“ problems. This was felt tolls satisfactory as
(a) no subject subjectively seemed disappointed at his score, (b) this
performance left a rather wide goal to shoot at both in terms of further
learning and in terms of aspiration.

A similar study of the second task was attempted with the same two
groups. In this case 30 seconds seemed optimal, the mean performance

being 7.25 pins.

Experimental Procedure

General set. The major problem faced in any experiment of they type is
to motivate the subjects in such a way that success and failure in the
task.are rather seriously taken. That is in such.a way as to ego—involve
the subject. This was accomplished in several ways: (a) All subjects
were told that we were interested in the intellectual ability and.mechan-
ical ability of patients in v.1. Hospitals and that this information was
to go in their case files. Further it was to be used as a basis for
further improvements in the educational and shop programs in the hospitals,
(b) An air of excitement and of spurring them on was maintained by the
examiner during the practice session. Thus if a subject finished six
problems he was told, "That's wonderful but I'll bet you can do more than.
that, (c) Whenever possible time was called in the muddle of a problem
or as a pin was about to be placed in the hole. These were not counted

but the examiner used this to encourage the person: ”Gee, you almost got

M9

that one. Let's try to get it next time.“ (d) The Success Trials (vide

infra) were also used to motivate the subjects and to spur them on.

Instructions. The instructions to the task were designed to be as clear

 

as possible, particularly in regard to the level of aspiration, in order
to avoid the possibility of getting several interpretations of what was
required (Er. Gould (26) and Festinger (13)].

The instructions may be divided into two parts (1) instructions for
the learning series (trials 1-5), (2) instructions for the goal setting
series (trials 6—25). The instructions for the learning series are as
follows: "Psychologists have found that the ability to do arithmetic
problems is closely related to intelligence. So we may consider this a
test of intelligence like the one you did before (Wechsler). You are
going to be given some simple addition problems; you will find that they
are all very easy and if you were given plenty of time you.could finish
them all. But you will only get 20 seconds in which to do them.-- that's
not very long. I want you to get all the problems you do right because
wrong answers don't count in the score but I also want you to go as fast
as you can. Ready - Start o'. At the end of the trial he was told he
had done very well and encouraged to do better on the next trial.

The instructions for the second task were essentially the samm:
"Here is a test of mechanical ability. The object is to get the pin in
the hole using the tweezers. You.will note that if you hold the pin by
the middle (examiner demonstrates) it's pretty hard to get it in the
hole. If you pick it up by the end nearest you (examiner demonstrates)
the pin pops out of the tweezers, but if you pick it up by the end

farthest from.you (examiner demonstrates) it is pretty easy to put in.

50
Now you fill in the first 10 holes to get the feel of it. (Subject does

this). Now let's see how many pins you can put in in 30 seconds."

The necessity of the practice and demonstration stems from the fact
that the learning trials were designed not only as a basis for equating
the groups as to ability but also to provide the motivation necessary
for ego involvement.

The level of aspiration instructions were as follows: "We are go.
ing to do some more problems just like the ones we finished. You will
have the same amount of time to do them (20. or 30 seconds depending on
the task). Before you start them I want you to make a bid like in Pinochle
or Bridge. I want you to tell me how many problems you will finish next
time. (Examiner answered any questions which occurred at this point.)
Remember before you start I want you to tell me how many you are going
to do. I will not remind you of this, you are to remember to do it on
your own. New have you decided how many you will do next time? How
my?“ (Time taken to respond is recorded as reaction time). If no
response occurs after 1&5 seconds the S is again reminded that he is to

tell how many he will do.

Genera; Procedure. The testing situation can be divided into five sec-
tions, four of which are repeated. They are (l) C.A.S. of the Wechsler
Bellvue, (2) Five practice trials with the arithmetic problems, (3) Five
practice sessions with the Tweezer Dexterity Test, (14) Five success trials
with the arithmetic problems, (5) 15 success-failure trials with the arith-
metic problems (a failure trial followed by two success trials, repeated
five times), (6) Brief interview. This series is followed by a ten-min-
ute break after which sections 3-6 are repeated with the Tweezer Dexterity

Test. The total session averaged about 90 minutes.

51
Thg_§,é,§, The C.A.S. consists of the Comprehension, Arithmetic and
Similarities subtests of the Wechsler-Bellvue Intelligence Test. This
test was administered after the first few minutes which consisted of
attempts on the examiner's part to put the subject at ease, as well as
to ascertain age and education for all subjects and degree of contact
for the schizophrenic patients. The test was scored as the subjects
took it and an I.Q. was obtained within a few minutes after its admin»
istration. This was made necessary in order to determine if the sub-

ject was suitable for the rest of the testing.

The practice sessions. The subject was given five trials with the
arithmetic problems (each set of 30 problems being presented on one
sheet). This was followed by five practice sessions with the Tweezer
Dexterity Test. The aim here was to assign the subjects into one of
two subgroups. This was done on the basis of both the normative data,
as well as of the performance data. Thus, if individual A.was 22 years
of age, with 8 years of education and an I.Q. of 101, and if his mean
performance on Task 1 were 7.5 and on Task 2, 8.5 problems, the exam-
iner would endeavor to assign a person of the same age, educational and
intellectual level and performance ability to the other subgroup. It
is obvious that the same age, etc., etc. do not frequently occur in two
individuals. However, the examiner was fortunate enough to get fairly

close approximations.

The subgroups. 0n the basis of the two preceding steps the individuals
of both the control and schizophrenic groups were placed in either the

strong or mild failure subgroups. Originally the writer had.hoped to

52
use the empirically tested criteria of Steisel and Cohen (67) to deter-
mine strong and mild failure, namely: mild failure was equal to being
stopped within five or six problems of the stated goal, whereas strong
failure meant being stopped within 10 or 12 problems of the goal. Bosh
ever, the rather low level of performance of the subjects made this imp
possible. It would be difficult indeed to tell a person whose average
practice performance was around seven or eight problems in twenty seconds
that his time was up after he finished barely two problems. It would be
impossible to have him fall within 10 or 12 problems of his goal as he
more often than not set his goal between 8 and 10 in the first few trials.
Thus mild failure was arbitratily defined as coming within 10 percent
of one's goal, while strong failure was defined as coming within 50 per-
cent of one's goal. The analysis of the first few records convinced the
writer that the results seemed satisfactory (i.e. that there were some
noticeable differences in the goal setting behavior of the two subgroups
of both the control and schizophrenic subjects) and so this arrangement
was maintained. Sixteen schizophrenics were placed in the mild failure
group (Group I), sixteen controls in the mild failure group (Group II),
sixteen schizophrenics in the strong failure group (Group III), and six-

teen controls in the strong failure group (Group IV).

The success trials. After the individual was assigned to a subgroup he
was given the level of aspiration instructions (vide supra) and the exa
perhment proper began. The first five trials in both tasks were “success"
trials. That is time was not called until he had finished the number of
problems he said he would finish (reached his level of aspiration). The

purpose of these trials was to observe goal setting behavior in a

53
rewarding situation, and to obtain as a basis for comparison with trials
in a nonprewarding or frustrating situation, trials which were not con—

taminated by the memory of previous failure.

The success-failure series. The five success trials were followed by a
series of 15 trials which began with failure (either mild or strong).

Each success trial was followed by the following comment: "You did
very well, you made your bid, that's fine.“ I

Each mild failure trial was followed by: "I'm afraid you didn't
quite make it. Ion came within x problems of your bid. That's not too
bad but it's not too good either.“

The strong failure trials were followed by: "You did very poorly,
you only got x problems, that‘s x problems below your bid. That's not
good at all.” These remarks were ‘usually preceded by an expression of
surprise and disappointment on the part of'the examiner.

The purpose of these remarks was to reenforce the objective failure,
and thus produce a frustrating state of affairs.

The success-failure series was then followed by a brief questioning
period during which the following standard questions were asked: (1)
How did you.feel after you failed to do as many problems as you said
you.would? (2) How did you feel when you did as many problems as you

said you would?

Timing. The reaction time (or time required to state a level of aspira-
tion) was recorded. The time it took to do the particular number of
problems was also recorded. Since the examiner felt that despite the
difficulty in estimating time, there might be some suspiciousness con-

cerning the accuracy of his measurement of it, particularly with the

5%
strong failure schizophrenic group, the following method was used:

1. In all of the learning trials the stop watch was placed in
such a way that the subject could see it. Attention was called to it
hy the examiner by remarks such as: "Time's up, that's 20 or 30 seconds,
isn't it?” or "The stop watch never lies - I guess time's up." Etc.

2. During the fourth or fifth success trial the stop watch was
stopped at the expected time (20 or 30"), the examiner using a wrist
stop watch to record the actual time. Again the subjects' attention was
called to the stop watch. This mine ruse was used after the first fail-
ure trial and after the 3rd or Nth failure trials. Since no comments
were made either during the experiment or in the interviews, one might

assume that the method was successful.

Order gf_presentation. Since the examiner could find no advantage to

 

a counter balanced order of presenting the tasks (Ff. Lewin 93 El. (5D
and Gould (26)] it was not used. Specifically, since (a) the tasks
were presented as discontinuous by the examiner to avoid transfer ef-
fect, (b) the early levels of aspiration were based on the levels of
performance of the practice trials, there seemed to be no necessity

for counterbalancing the two tasks. Thus the arithmetic test was always

the first test.

Summagz. The objective of this study was to observe the effect of fail-
ure on goal setting behavior. In order to do this efficiently it was
felt that (a) degree of failure should be varied (b) failure and success
trials should be separated. Further the groups should be equated for

age, education and intelligence and performance on tasks.

55
The task itself should be one which could be relatively ego-invol-

ving, and which would fit into a scheme of controlled success and fail-
ure.
To an extent these objectives would seem.to have been fulfilled.

(A more unified concept of the experimental procedure may be obtained

from Table 2 which follows).

56

.uneooom v

.Hsom «an no pseoaom om manna: neaoo poenadm t Aoanawch wdoapmv hm
.Hoow use so sooosod oH sagas: moaoo ooonpam u Aoasaama eaazv a: o

.eaoo- on ens. eon on msoaoosd so Hogans one eoooadaoo eon soonest on» Hanna .aa» wqaaaoo son
an soaocafiams no Ho>oa ego e» Hmsve eons new eoqdaaomaem no Ho>oa on» madam» awesome on» ma 9

.eaa» ecsooom on no cm can a“ oopoamaoo can on eaeapoam hose tom nae» as:
soonpsu one .aooaesue an confide haepsasooc as: Azom ac omv eaaa edema» wsauasea on» ea c

 

m m hm m m hm m m mm m m mm m m hm .odm .namen

m m hm m m hm m m hm m m mm m m mm .ofim .namen

m m h: m m ME m m m2 m m as m m m2 .odm .nhmen

m m as m m h: m m as m m as m m MS .odm .aameq

mdo

m<o

m<o

mdo

easafish mnoaam
Hoaauoo
933cm $3.3m
caaeammouamcm

onflafimh_cawz
Heapsoo

easaaeh wag:
caneamnoaamom

 

mm em mm mm mm om mm mm am so ma so no. emu odd ooa.m sm-o
. H o a a a

 

 

mmbomm Adezmzummmxm mma mom mumda mo HoszGMm AdMDnmoomm

m sense.

F INDINGS
W 9.1; $129.13.

As has been mentioned before our interest was directed to the ef.
fects of success and failure on several variables, 31;, Level of Aspir-
ation, Discrepancy Score, Efficiency, Number of Errors, Reaction Time,
and Responses to Questions. With the exception of the last mentioned
variable, the data were statistically analyzed in the following manner:
(a) the mean score for each individual was obtained for the success
trials (trials 6—10), as well as for the immediate post-failure trials
(trials 12, 15, 18, 21 and 2M) and the second post-failure trials (trials
13, 16, 19, 22, 25). It will be noted that these post-failure trials
are the success trials ihich follow the failure conditions. Immediate
and second post—failure trials were separately analyzed in order to note
the possible existence of a delayed reaction to failure. (b) A group
mean was then calculated, i.e. the mean of the means of the success trials
etc. was computed for each of the two schizophrenic groups and for their
respective control groups. (c) Fisher's t formula for small samples (30)
was then utilized for the following comparisons:

1. Intragroup comparisons
a. Differences of means between success and immediate post-failure

trials,
b. Differences of means between success and second post-failure

trials.

c. Difference of means between immediate and second post-failure.

58
2. Intergroup comparisons

a. Difference of means between schizophrenics in the mild failure
condition and their controls for all three types of trials (success, im-
mediate and second post failures).

b. Difference of means between schizophrenics in the strong failure
condition and their controls for all three types of trials.

c. Difference of means between the two schizophrenic groups for
the three types of trials.

d. Difference of means between the two control groups on each of
the three trial types.

All of these comparisons were made for each task (Task A, arithmetic
problems -- Task B, peg board) with the exception of mean number of errors
which was only computed for Task.A, as errors were not scored on the
second task.

Since our hypotheses had ”a directional character” the one tailed

t test was utilized. (RM)..
Results g£_Level g£_Aspiration

The level of aspiration represents the subject's own estimate of
what he will do on a trial which he is about to start. In this study
it was given in the number of problems the subject felt that he could
solve or the number of pins he could place within the given time limit.

The hypotheses concerning this variable were:

(a) that the schizophrenic group would give a higher level of as.

piration following failure than would the controls,

59

(b) that there would be a higher level of aspiration with this
group folowing failure than there was in success,

(c) that the schizophrenic subjects given strong failure would
give still higher levels of aspiration than those in the mild failure
condition. I

The data in Tables 3, 3a and 3b substantiate these hypotheses.

The difference between the two schizophrenic groups and their controls
reach a statistically significant level after failure though none exists
in the success trials.

When the schizophrenic mild failure group is compared to the schizo-
phrenic strong failure group a statistically significant difference is
also obtained.

One additional finding which was not predicted stands out very
clearly. The control group in the strong failure situation shows a
higher level of aspiration than does the control group given mild fail-
ure. Further the comparison of the success trials with Second Post-
Failure trials for the former group also shows a significantly higher
level of aspiration after failure.

It would seem that though the control groups are less sensitive to
failure than the schizophrenic groups, strong failure does tend to yield
higher levels of aspiration.

This seems consistent with the postulation of a threshold for frus-
tration. In those subjects who are markedly sensitive to frustration
(the schizophrenics) a small amount of it will lead to unrealistic aspi-
rations, whereas in the case of those subjects who are not "reaction

sensitive" it takes a greater degree of frustration to produce this

TABLE 3

MEAN LEVEL OF ASPIRATION FOR TWO SCHIZOPmENIC AND T170 CONTROL
GROUPS IN CONDITIONS OF SUCCESS AND FAILURE ON TWO TASKS

 

 

 

Immediate Second
Succe s s Po 3 t-Failure Po st-Fai lure
Groups Task Trials Trials Trials
M S.D. M S.D. M S.D.
Schizophrenic A 11.81 2.93 13.75 2.27 114.75 2.17
Mild Failure
B 13.75 2.29 15.50 2.82 16.75 2.19
Control 1 12.uu 1.67 11.9u 1.53 12.uu 1.87
Mild Failure
B 1h.25 1.76 lu.00 1.9a 1h.75 1.88
Schizophrenic 1 12.00 2.67 15.56 2.05 ' 17.50 2.10
Strong Failure
B 1h.56 2.18 17.25 3.u9 19.50 2.1M
Control A 12.uu 1.uu 13.12 1,u2 13.62 1.87
Strong Failure
B lu.25 1.60 15.25 2.56 16.00 1.80

 

TABLE 3a

61

DIFFERENCES BETWEEN TYPES OF TRIALS ON MEAN LEVEL OF ASPIRATION
FOR TWO SCHIZOPHRENIC AND TWO CONTROL GROUPS IN CONDITIONS
OF SUCCESS AND FAILURE ON TWO TASKS

 

21222211221.
'Immediate Second
Success Post-Failure Post-Failure
Groups Task Trials Trials Trials
Diff. _ t Diff. t Diff. t
Schizophrenic A .61 .72 1.81 2.73‘* 2.31 3.27*‘
Mild Failure and
Control Mild
Failure B .50 .70 1.50 2.12"I 2.00 2.39“
Schizophrenic A .uu .7u a.hh u.06*- 3.88 5.6Mfl
Strong Failure
and Control
Strong Failure B .31 .65 2.00 2.93** 3.50 3.33MI
Schizophrenic 1 .19 .31 1.81 2.140... 2.75 3.76*¢
Mild Failure and
Schizophrenic
Strong Failure B .81 1.03 1.75 2.28* 2.75 2.5N**
Control Mild A O O 1.18 1.96* 1.18 1.82.
Failure and
Control Strong
Failure B O O 1.25 1.89* 1.25 1.65'

 

’ Significant at the 5 per cent level or less.
** Significant at the l per cent level or less.

62

 

 

.enoa ac Hoaoa acme pom H can as psseflmasmHm es
.umoa no Ho>ma ease non m on» as assofimasmam e
:o.~ ma. .0m.m mm. ao.a ma. ~m.H am. m oneness
ndmom unooom
aw. om. ..o~.m am.” am. om. Hm.a ma. 4 sec occaeoasH
..Om.m mk.a ..m~.m :m.: as. aw. ..mm.m oo.m m confines
npuom uncoom
.mo.m wa.a ..Hm.m om.m o o ..zm.m :m.m 4 sec ooooosm
NM.H oo.a ..o~.m wc.m Hm. mm. .mo.m ma.a m essence
.c.om .coaeossa
am.a ma. ..zm.s hm.m om. om. .mo.~ :m.a 4 ecu ..ooosm
a .ccan » .esan a .sman » .ccan
oneness eschew cheeses mcohcm cheeses can: cheeses can: shoe Hosea co case
Hospcoo owscagaoufinom Honpcoo owcoanmcuumom
scam

 

 

mumde 039 20 QMDAH<M 924 mmfioobm mo szHBanoo zH mmbomo Rowszoo

039 924 onmmmmouHmom 038 mom zOHsdemmd ho qm>mq 24M: 2H mmOzmmmamHn

pm amass

63
result and even then it is not of the asme magnitude as that of the

schizophrenic subjects who also experienced strong failure.

Results of Score

 

The D Score is the discrepancy between the previous level of achieve-
ment and the present level of aspiration. It is positive if the subject
states he will do more on the nextirial than he did on the last and nega—
tive if he states he will do less.

The hypotheses concerning this variable were:

(a) that the schizophrenic subjects would give a higher positive D
score than the controls after failure,

(b) that the schizophrenic subjects would give higher positive D
scores after failure than after success,

(c) that they would give higher D scores after strong failure than
after mild failure. A

These hypotheses are confirmed by the data in Tables N, ha and Nb.
Though no statistically significant difference between the schizophrenics
and their respective controls appear when success trials are compared,

a great difference is noted between them.on both post-failure trials.
The schizophrenics also differ, the strong failure group giving the higher
D score.

The confirmation of the hypotheses concerning level of aspiration

and discrepancy scores lead us to accept the first major hypothesis, 115:

After failure the schizophrenic subject will be less influenced by his

post performance in goal setting than will the control group.

MEAN DISCREPANCY SCORE FOR TWO SCHIZOPHRENIC AND TWO CONTROL
GROUPS IN SUCCESS AND POST-FAILURE TRIALS ON TWO TASKS

1

TABLE h

6h

 

Immediate Second
Success Post-Failure Post-Failure
Groups Task Trials Trials Trials
M S . D. M S. D. M S. D.
Schizophrenic A .69 .29 .9M .49 .9M .M5
Mild.Failure
B .62 .2“ 1.31 .58 1.12 .59
Control A .56 .13 .56 .26 .50 .28
Mild Failure 7
B .69 .25 .62 .33 .69 .35
Schizophrenic A. .62 .33 2.69 1.02 2.18 .97
Strong Failure .
B .69 .25 2.62 .79 2.31 .95
Control A .69 .18 .69 .“O .62 .34
Strong Failure
3 .69 .2h .81 .MA .81 .55

 

TABLE 4a

65

DIFFERENCES BETWEEN TYPES OF TRIALS 0N MEAN DISCREPANCY SCORES
FOR TWO SCHIZOPHRENIC AND TWO CONTROL GROUPS IN CONDITIONS
OF SUCCESS AND FAILURE ON TWO TRIALS

 

 

 

Type of Trial
Immediate Second
Success Post-Failure Post-Failure
Groups Task Trials Trials Trials
Diff. t Diff. t Diff. t
Schizophrenic A .13 1.64 .38 2.75** .44 3.33"”I
Mild.Failure
and Control
{ild Failure 13 .07 1.19 .69 4.11" .144 2.561-
Schizophrenic A .07 .74 2.00 7.30** 1.50 6.05‘*
Strong Failure
and Control
Strong Failure B O O 1.81 8.04** 1.50 5.47MI
Schizophrenic A .07 .64 1.75 6.18“I 1.24 4.59"“.I
Mild Failure and
Schizophrenic
Strong Failure B .07 1.19 1.31 5.39** 1.19 4.23""I
Control Mild. A .13 2.36*‘ .13 1.09 .12 1.08
Failure and
Control Strong
Failure B O O .19 1.39 .12 .74

 

* Significant at the 5 per cent level or less.
** Significant at the 1 per cent level or less.

66

.umca ac Ho>mH «coo non H on» as unsoamwsMam as
.emoa ac Hopoa pace sea m on» as acsoachme s

 

 

o o oo.H am. am. 80. mm. Hm. m cheeses
lanom dsooom
mm. No. .No.m aa. an. mo. 0 o q ecu openeoaaa
so. an. ..mm.h mm.a o o ..mm.m mm. m oneness
ipmom osooom
ma. No. ..No.c mm.a mo. 00. .Nm.a mm. < soc moooosm
mm. ma. ..m:.m em.” om. mo. ..mm.: as. m cheeses
leach upswuoEEH
o o ..ma.~ ao.m o o .Hw.a mm. 4 sec moooocm
» .caan p .ccan » .aean » .ccaa

cheeses mooocm cheeses scoocm essence can: oneness sass smog Hence so case

Hoaacco cascanaoaamom Hoapaoo owsoamacuamom , .

mdoam

 

I"
’I” I;

mxmde 0:9 20 mmbHH4h 924 mmmoobm ho moneanoo 2H mmbomo
Howezoo 039 924 UHZEMmMONHmom 039 Mom mmmoom NOZHNMMOmHQ 24m: 2H mmozmmmmmHm

p: mamas

67
One might see this as related to loss of contact with reality. No
longer can past experience serve as a guide to future action; the pain-
ful past is in effect denied by the continuance of high aspirations after
failure. This is not hopefulness but an obstinate refusal to face facts.
We note that our controls exhibit a somewhat higher level of aspiration
after strong failure and conclude that these differences are a matter
of degree rather than of kind, that under frustration the most "normal"
human being will also lose contact as it were and act in a most unreal-

istic manner.
Results on Efficiency

Efficiency is calculated by dividing the number of units into the
time taken, i.e. it represents the number of seconds to either solve an
arithmetic problem or to put a peg into the peg board.

The mincrhypotheses concerning efficiency were:

(a) that the schizophrenic subjects would.show a greater drop in
efficiency following failure than would the controls,

(b) that there would be a greater drop in efficiency after failure
than after success trials for this group,

(c) that schizophrenics receiving strong failure would be less ef-
ficient than those receiving mild failure.

Tables 5, 5a and 5b indicate that these hypotheses are confirmed.
There is indeed a drop in efficient performance after failure among the
schizophrenic subjects, more severe for the strong failure than for the
mild failure subgroup. However, a drop in efficiency is also noted in

the strong failure control group in Task A (Table 5b).

TABLE 5

MEAN EFFICIENCY SCORE FOR TWO SCHIZOPHRENIC AND TWO CONTROL
GROUPS IN SUCCESS AND POST-FAILURE TRIALS ON TWO TASKS

m T ‘vu
h— h f --

 

 

 

Immediate Second
Success Post—Failure Post-Failure
Groups Task Trials Trials Trials
M S.D. M S.D. M S.D.
Schizophrenic A 1.94 .52 2.44 .59 2.50 .57
Mild Failure -
s 3.50 .49 n.00 .70 4.06 .66
Control 1 1.87 .51 1.87 .54 1.94 .51

Mild Failure
3 3.56 .M9 3.37 .59 3.uu .h6

schizophrenic A 2.12 .56 2.87 .55 2.94 .58
Strong Failure
B 3.62 .60 n.5o .65 4.56 .65

Control A 1.87 .45 2.19 .52 2.19 .48
Strong Failure _
B 3.56 .44 3.69 .48 3.81 .u7

 

59

TABLE 5a

DIFFERENCES BETWEEN TYPES OF TRIALS 0N MEAN EFFICIENCY SCORE FOR
TWO SCRIZOPHRENIC AND TWO CONTROL GROUPS IN CONDITIONS OF
SUCCESS AND FAILURE ON TWO TASKS

w‘w‘ w "vw *- n ‘ fl v—‘
‘

innate-iii

 

 

 

Immediate Second
Success Post-Failure Post-Failure
Groups Task Trials Trials Trials
Diff. t Diff. t Diff. t

Schizophrenic A .07 .43 .53 2.72“ .53 3.42MI
Mild.Failure and
Control Mild
Failure B .06 .30 .63 2.85** .63 2.90“
Schizophrenic A .25 1.43 .68 3.49** .75 3.39*'
Strong Failure
and Control
Strong Failure B .06 .32 .81 4.05“I .75 3.81**
Schizophrenic A .18 .96 .43 2.08"l .44 2.09'
Mild.Failure and
Schizophrenic
Strong Failure B .12 .57 .50 2.11"l .49 2.17‘
Control Mild A 0 0 .32 1.76s .26 1.53
Failure and
Control Strong
Failure B .03 .18 .32 1.76* .37 1.98‘

 

* Significant at the 5 per cent level or less.
‘* Significant at the 1 per cent level or less.

7O

.nncH no Hoaoa ammo sea A on» as vacoguasMam as
.uuoH ac Ho>oa peso arm m on» as ansoamacwam e

 

mm. ma. Hm. mo. 0:. 20. mm. No. m cheeses

nanom cnooom

o 0 mm. No. an. we. on. om. 4 sec occaeossa

mm.a mm. ..mm.: :m. mm. as. ..om.m am. m chanson

tamom csooom

.mm.a mm. ..mm.m mm. mm. 80. ..mm.m mm. 4 sec .oooocm

am. ma. ..ma.m mm. mm. ma. .ma.m om. m confines

Iauom oucwcoaaH

.mm.a mm. ..mm.: ma. 0 o ..mm.m On. 4 sec ..oooem
a .ccan c .aaan » .csan c .acan

oneness enoncm confines sconce oneness can: chanson can: sore Hosea co case
Hoapnoo oasoanmouwnom Hoapnoo cascannoewnom

muons

 

H

mMmde 029 20 HmDHHdfi 924 mmmoobm 90 m2099992oo 2H mmbomw
9098200 029 92¢ 09292290N920m 039 mom mfimoom Mo2mHlomfi 2492 29 m9029m99299

pm mam4a

71
Further the mild failure control group is apparently more efficient
after failure than is the strong failure group on both tasks despite
the fact that there is little difference in their efficiency in the suc-
cess trials. Thus we again emphasize the fact that failure and frus-
tration disrupt all behavior to some extent and that the more sensitive

are most disrupted.

Results on Number of Errors

“mum

The error score is merely the number of incorrect addition problems.
An error score was not feasible for the second task since only the num»
ber of page put in counted toward the total score; whereas on the arith—
metic task all problems finished were assumed correct and entered into
the total score, i.e. as goal achievement.

Our predictions concerning errors were:

(a) that the schizophrenic groups would make more errors than the
control group after failure,

(b) that they would make more errors after strong failure than
after mild failure.

The means and t values for this variable may be found in Tables 6,
6a and 6b.

In general our hypothese are confinmed. The schizophrenic groups
do make significantly more errors than their controls and the schizo-
phrenic strong failure group does make more than the mild failure group.
There is no difference in number of errors between the two control groups.
(cf. Table 6a). In the intra-group comparisons (Table 6b) we note sig-
nificant differences between success trials and the post failure trials

for both schizophrenic groups but none for the control groups.

72

TABLE 6

MEAN NUMBER OF ERRORS FOR TWO SCHIZOPHRENIC AND TWO CONTROL GROUPS
IN SUCCESS AND POST-FAILURE TRIALS ON ONE TASK

 

.9 a:

 

 

Immediate Second
Success Post-Failure Post-Failure
Group Trials Trials Trials
' M S.D. M S.D. M S.D.
Schizophrenic
Mild Failure .25 .140 1.31 1.21 1.37 1.15
Control
Mild Failure .25 .58 .31 .66 .37 .81
Sdhizophrenic
Strong Failure .31 .47 2.75 2.53 2.87 2.40
Control

Strong Failure .25 .49 .62 .72 .69 1.09

 

73

TABLE 6a

DIFFERENCES BETWEEN TYPES OF TRIALS 0N MEAN NUMBER OF ERRORS FOR
TWO SCHIZOPHRENIC AND TWO CONTROL GROUPS IN CONDITIONS
OF SUCCESS AND FAILURE ON ONE TASK

 

Immediate Second
Success Post-Failure Post-Failure
Groups Trials Trials Trials
Diff. t Diff. t Diff. t
schizophrenic Mild o 0 .96 2.71** 1.00 2.98"
Failure and Control
Mild Failure
Schizophrenic Strong .06 .37 2.13 3.03*‘ 2.18 3.37“
Failure and Control
Strong Failure
Schizophrenic Mild .06 .39 1.44 2.06* 1.50 2.30:
Failure and-
Sdhizophrenic Strong
Failure
Control Mild. Failure 0 O 006 025 032 096
and Control Strong
Failure

 

‘ Significant at the 5 per cent level or less.
** Significant at the l per cent level or less.

74

.neoa 9o Ho>oH pace sea a on» as assoa99cM9m so
ecoo non m on» he seconaanmaw .

.anoH ac Hobca

 

Hm.

mm.

am.

a

90.

:z.

9m.

.9999

90.

ssm~.m

ssHm.m

»

mo.

NH.H

00.9

.9999

cauaamm

tench csoocm
0nd camacoEBH

unseacm

lumen uaooom
one uncoosm

. oaaaash
:pnom cacauanH
one onoooam

 

endeash mcoapm
Hospnoo

:a. ma. mm. so.
..am.: mm.m em. ma.
..mm.m e:.m Hm. mo.

9 .9999 a .9999
cadaacm macapm canaash 0992
cascamnouwsom Hoapcoo
macaw

oa599s9 c995
camcannoewmom

Henna so ease

 

 

pm 99949

Mm¢9 920 20 9909949 924 mm9000m 90 m209999200 29 mmbomw
9099200 028 924 09299290N990m 029 909 mmommm 90 999202 2492 29 m902M999999

75

Thus our second major hypothesis, which states in effect that fail-
ure will disrupt the performance of sdhizophrenics to a greater extent
than it will the performance of "normals" seems to be confirmed. It is
interesting to note that while the "normals" in strong failure do ex-
hibit an increase in the time taken to do the addition problems (Task A)
they do not show any greater number of errors in their works The
schizophrenics even in mild failure not only are slower but are less
accurate.

One might speculate about this difference in terms of recognition
of failure on the part of the normal subjects who then attempt to com.
pensate by striving for accuracy at the expense of speed. The schizo-
phrenics, on the other hand, are perhaps thrown into a sort of panic
state where they can neither move faster nor produce well. This, of
course, must remain speculation.

The third major hypothesis concerns the attempt to withdraw from
a failure situation on the part of our schizophrenic subjects after fail-

ure.
Results g£_Reaction Time

Reaction time is defined as the time (in seconds) which the subject
takes to decide on.his next bid (statement of level of aspiration). Our
minor hypothesesconcerning it were:

(a) that the schizophrenic subjects would show a longer reaction
time after failure than in success,

(b) that the control groups would not show this difference,

(c) that a difference would exist between the schizophrenic mild

76
failure group and the schizophrenic strong failure group, the strong
failure group giving the longer reaction time,

(d) that no such differences would exist between the two control
groups.

The means and standard deviation for this variable are shown in
Table 7, and the t ratios in Tables 7a and 7b.

We note that the schizophrenic mild failure group gives a signif-
icantly longer reaction time after failure than it does in the success
trials. This applies to both immediate and second post failures except
in Task B where the comparison between success and second post failure
is not significant. There is not statistically significant increase or
decrease in reaction time between the two post failure trials.

The control mild failure group on the other hand shows no signif-
icant differences between their reaction time in success trials and
that in post failure trials.

The schizophrenic strong failure groups shows highly significant
increases in reaction time after failure, on both tasks. As with the
other schizophrenic group no differences of significance are obtained
between the two post failure trials.

The control group does show an increase in reaction time after
strong failure.

This failure to state the next level of aspiration following fail-
ure has been observed by many experimenters. Escalona (10) and Rotter
(62) have interpreted it as withdrawal behavior while fliller (57) bases
his operational definition of "conflict" on this reaction, a concept

which he interprets in much the same way as do Rotter and Escalona.

77

TABLE 7

MEAN REACTION TIME (in seconds) FOR TWO SCHIZOPHRENIC AND TWO CONTROL
GROUPS IN SUCCESS AND POST-FAILURE TRIALS ON TWO TASKS

 

 

 

Immediate Second
Success Post-Failure Post-Failure
Groups Task Trial 3 Trial 8 Trial 3
M S.D. M S.D. M S.D.
schizophrenic A 7.12 1.89 8.75 2.18 8.94 2.14
Mild Failure
B 8.94 1.24 9.75 1.32 9.50 1.27
Control A 6.50 1.20 7.00 .91 7.00 .93
Mild Failure
B 8.50 .92 8.62 .99 8.75 .92
Schizophrenic A 7.50 1.70 10.25 1.76 10.44 1.88
Strong Failure
B 8.94 1.73 10.94 1.91 - 10.25 1.84
Control A 6.75 1.54 7.75 1.46 7.75 1.44
Strong Failure
B 8.62 1.38 9.44 1.45 9.50 1.38

 

78

TABLE 7a

DIFFERENCES BETWEEN TYPES OF TRIALS ON MEAN REACTION TIME FOR TN
SCHIZOPHRENIC AND TWO CONTROL GROUPS IN CONDITIONS OF
SUCCESS AND FAILURE ON TWO TASKS

 

 

 

2222.2£.22121.
Immediate Second
Success Post-Failure Post—Failure
Groups Task Trials Trials Trials
Diff. t Diff. t Diff. t

Schizophrenic A .62 1.09 1.75 2.97~* 1.94 3.329.
Mild Failure

and Control
Mild Failure B .44 1.05 1.13 2.71" .75 2.15:
Schizophrenic A .85 1.40 2.50 4.46** 1.13 2.71"

Strong Failure
and Control

Strong Failure B .34 .56 1.50 2.55.. .75 2.34:.
schizophrenic A .48 .77 1.50 2.18* 1.50 2.15.
Mild Failure and

Schizophrenic

Strong Failure B 0 O 1.18 2012‘ 075 1087.
Control Mild A. .25 .62 .75 1.69* .75 1.70.

Failure and
Control Strong
Failure B .12 .25 .82 1.86‘ .85 2-33"

 

* Significant at the 5 per cent level or less.
** Significant at the 1 per cent level or less.

79

.nuca no ao>ca pace sea A on» as pqc0999sm9m es
.nmoH no Hobca ammo pom m on» as aqd0999c09m s

 

 

om. mo. wo.H me. on. ma. an. an. m cheeses

Ipmom cnooom

o 0 mm. me. o 0 am. ma. 4 and eeoaeossH

mw.a an. .oa.m Hm.a mo. mm. .ma.a mm. m oneness

lauom csooom

mm.n o.a ..me.e em.m , mm.a om. ..am.m mu.a 4 ans enooecm

om.a mm. ..ha.m oo.m on. ma. ..mh.m aw. m oaeaanateeom

opmfiooaaH

om.a o.a ..mz.e ma.m mm.a cm. .em.m mm.a 4 one ..ooeem
a .uaaa n .9999 e .muan e .9999

caeaaoaaueonem essence unenem cheeses seas onsaaea can: More areas no case
Homecoo cascannouasom Hosanoo owcoammoeamom

macaw

 

 

m9m<a 029 20 9999949 924 mmfioobm 90 m209999200 2H m99090
9099200 099 924 09299990N990m 029 909 9299 20990499 2d92 2H m9029999999

pa mam4a

80
Thus some withdrawal is noted in the "normals" under conditions of
strong failure. The schizophrenic subjects,however, show this tendency

even in mild failure and mush more markedly in strong failure.

Results gg_guestions

A further indication of an attempt to withdraw from the conscious-
ness of failure may be drawn from.the classification of the answers
given to the two questions asked at the end of the testing session.
They were: (a) "How did you feel after you failed to do as many prob-
lems (put in as many pegs) as you said you would?” (b) "How did you
feel when you did as many problems (put in as many pegs) as you said
you.would?"

Our minor hypotheses state that the schizophrenic patients will
attempt to deny failure more than the controls and that they will tend
to deny it more often after strong failure than after mild failure.

Of our 32 schizophrenic subjects, 20 or 62.5 per cent answered
that they did not feel one way or another about it -—- "It didn't bother
me at all.” "Failure! There was no failure, I wanted to make sure I
was right.‘I Only 9 or 28.1 per cent of the controls responded in this
manner.

Eleven of the 20 schizophrenics denying failure were in the strong
failure group, the other nine being in the mild failure group, while 5
of the 9 controls were in the strong failure group.

The responses above are further clarified by the fact that only 4
or 12.5 per cent of the schizophrenic subjects denied feelings of suc-

cess, while six of the controls, or 18.7 per cent, were similarly disin-

terested.

81

The denial of failure then does not seem to result from a general
lack of interest in the situation but rather from a need to perhaps for—
get about it.

Thus, our third major hypothesis is confirmed. It is interesting
to note that while many of the normals in strong failure withdrew from
the effect of failure in a more or less inconspicuous manner, by delay.
ing the statement of the level of aspiration, they could nonetheless
face the fact when it was directly thrust on them by the questions asked.
The schizophrenic subjects could not face the fact at all and talked as
if it did not matter at all, though succeeding did.

In general than we may state that schizophrenic subjects are quite
”reaction sensitive to failure" and that as a possible result of this
they tend to become more unrealistic and less efficient when faced with

it.

DISCUSSION

It was assumed in this study that schizophrenic patients, as a
function of their "reaction sensitivity" to failure would behave more
defensively and less efficiently after failure than a control group of
equated non-psychiatric subjects. In general these assumptions seem

to have been borne out.

Comparison tg_0ther Scizophrenic Group§_

 

These results agree essentially with those of Kyle (49) and Bowman
(5) in that the schizophrenic groups show a higher level of aspiration
and a higher deviation score than the control groups. Perhaps this
study has tended to clarify their results in that it demonstrated that
both of these variables show this trend after failure but not in suc-
cess, and also that in a situation of differential failure the schizo-
phrenics who are exposed to strong failure react in a more exaggerated
fashion than those exposed to the mild failure conditions. This, in
spite of the fact that each failure trial is followed by two success
trials.

In terms of these two variables we disagree with Miller's (57)
findings of little difference in level of aspiration and discrepancy
score between schizophrenics and normals after failure, though this
difference may be the result of the use of different populations. How-
ever, we do find as he did that reaction time increases after failure.
Again by showing that this applies only after failure and that it is a

function of degree of failure, Miller's results seem to be clarified.

53
The other two variables, efficiency and errors, have not been used in
level of aspiration studies with schizophrenics. However, Webb (69)
found that in a conceptual task those schizophrenics who had been told
they had done poorly on a pretest Showed greatest imprecision and tan-
gentiality on the post test while those who were not so discouraged did

significantly better on a post test.
Comparison Between Schizophrenics and Controls

Since this study involved two degrees of failure, the results ob—
tained with the control groups cannot be directly compared to those
obtained by Miller (57), Kyle (49) or Bowman (5). In general the con-
trol group who experienced mild failure reacts similarly to their con—
trols, i.e. their level of aspiration is not significantly higher in
failure than in success, their discrepancy scores are not higher, nor
is their reaction time longer. The controls in the strong failure sit—
nation, however, more closely resemble Sears (65) highly frustrated
group of normals in terms of their comparatively unrealistic and inef.
ficient behavior after failure. Though their behavior was disrupted.it
was not as disturbed as that of the two schizophrenic groups.

Thus, though they do tend to give higher levels of aspiration after
failure, they do not completely lose sight of their past performancetas
dilthe schizophrenics. Therefore, their discrepancy scores after fail—
ure are not significantly different from those in the success trials,
as are the D scores of the schizophrenics. It is almost as if the con-
trols hope for better results, after failure, but keep the failure in

mind. They take the probability of another failure into account, but

84

are not so depressed by past failure that they are content to stay at
the same level as before or even retreat to a lower level.

The schizophrenics seem to act as if they had never failed; they
continue to "go upward and onward" despite the very obvious fact that
they have not previously been able to reach a goal more modest than
the one they presently aspire to. Here one may speculate that the wish
reigns supreme unhampered by judgment.

In terms of efficiency of work, the schizophrenics not only takes
longer time to do the same amount of work after failure but they do it
less accurately. The controls in strong failure are also a little
stunned by failure but though their speed decreases, accuracy remains
at about the same level as in success. Their work is not as disrupted
by failure as is that of the schizophrenics. Further, the schizophrenics
show a greater "burnt child" reaction after failure than do the strong)
failure controls, They not only withdraw from the situation by taking
a longer time to get back into the task, but they deny subjective feel-
ings of failure.

These differences can be accounted for in terms of reaction sensi-
tivity. The schizophrenics are apparently quite sensitive to failure
even of the mildest sort. They have a minimal tolerance for frustration.
Failure leads to unrealistic and inefficient behavior of the grossest
sort.

The controls seem to be better able "to take it," but even they
have a threshold beyond which the effects of failure begin to tell. In
effect a continuum of frustration tolerance or tolerance of failure seems

to be demonstrated. The schizophrenic group in the strong failure

85
situation is the mast disrupted, the schizophrenics given mild failure
follow, then the controls given strong failure, and finally the control
group who experience mild failure.

Our findings are therefore not specific to schizophrenics; rather
these results seem to occur whenever frustration goes beyond the bounds
of endurance.

The neurotics studied by Miller (57), Himmelweit (35) and Eysenck
and Himmelweit (11) also behave in a rather disrupted.manner after fail-
ure. Perhaps they would fit in between the schizophrenics and the con-
trols on the proposed continuum. Hewever, this must be ascertained by

future researdh.

Implications

 

The findings of this study lend some weight to Jenkin's hypothesis
"that schizophrenia is a progressive maladaptation resulting from frus-
tration beyond the tolerance of the patient," (43, p. 256), as well as
to Cameron's theory of "reaction sensitivity"to frustration on the part
of schizophrenics (6).

What does this imply? The work of Sears (65), Katz (46) and Mar-
quart (56), as well as the results obtained from the control strong fail-
ure group indicate that continued failure to achieve one's goal does re-
sult in attempts to neutralize the experience of failure by various with—
drawal medhanisms. Jenkins (43) cites numerous studies which indicate
that schizophrenic patients have experienced severe frustration within
the family group at an early age. He feels that because of this they

are ill prepared for the frustrations of every day life and thus withdraw

86
from the world of reality. Whether schizophrenia can be entirely exa
plained on the basis of this hypothesis must remain a moot question,
awaiting further researdh. However, it does seem likely that reaction
sensitivity to failure, based perhaps on childhood experiences, may be
an etiological factor not only in schizophrenia but also in theinuroses.

Perhaps it is aspiring too high to hope for answers to questions
regarding the kind and amounts of frustration needed to produce emotional
illness. But certainly further studies of the effects of frustration,
especially longitudinal ones starting in early childhood, are to be
desired.

Further research along the lines proposed by this study might yield
more evidence of a continuum from mild to severe emotional disturbances.

Research on the rexersibility of this great sensitivity to failure
may also prove fruitful: Can a series of continuous successes, for exa
anple, "retrain" the individual to eventually absorb some frustration
without headlong retreat?

Whatever the answers, it seems certain that the area of goal set-
ting bears a great deal of study both by laboratory methods and by

broader sociological procedures.

SUMMARY AND CONCLUSIONS

Two groups of sixteen patients, diagnosed as mixed schizophrenics,
and two equated groups of patients at a general medical and surgical
hospital were given twenty trials on two tasks (addition problems,
tweezer dexterity) involving controlled success and controlled failure
on two levels of intensity. The following variables were studied: (a)
level of aspiration, (b) discrepancy score, (c) efficiency, (d) number
of errors, (e) reaction time.) All subjects were also asked for their
reactions to achievement and non-achievement of their stated goal.

Results obtained within and between groups by the use of the one
tailed t test indicate that (a) both schizophrenic groups show a
higher level of aspiration, a higher discrepancy score, lowered effic-
iency, greater reaction time, and more errors after failure than they
did after a series of success trials, (b) the schizophrenic groups
showed a significant difference in the above directions when compared
to their "normal" controls, (c) the schizophrenics in strong failure
conditions showed a significant difference in the above direction When
compared to the sshizophrenic mild failure group, (d) the control strap;
failure group showed higher level of aspiration, longer reaction time
and lowered efficiency after failure, (e) the control mild failure did
not react differently after failure than before. Analysis of the answers
to the questions which attempted to elicit verbal reactions to success
and failure reveal that the schizophrenic subjects will more often deny
any feeling of failure than will the controls, though they admit feelings

of success almost as frequently.

88

These data are interpreted in accordance with thelypotheses of
Cameron (6) and Jenkins (#3) regarding reaction sensitivity to failure.
The fact that the normal controls also show some effects of post-failure
frustration is seen as essentially consistent with these hypotheses.

In general then we may state that failure even when cushioned with
success, has a severe effect on schizophrenics. Further studies using
variable failure with other schizophrenic groups and with neurotic
groups should be attempted, as should studies involving retraining of
schizophrenics after failure by giving massed success trials.

Though the literature is filled with level of aspiration studies
there are still relatively few which attempt to relate goal setting to
mental disease. It seems to the writer that further investigations seem
warranted, as they may indeed add to our meager knowledge of psychody.

namics.

1.

9.

10.

ll.

12.

13.

BIBLIOGRAPHY

Abraham, K. The psycho-sexual differences between hysteria and
dementia praecix,in Abraham, K. Selected Papers gg_Psychoanalysis.

' London: The Hogarth Press. 19M8.

Alexander, F. ng.égg.g£_Unreason. New York: J. B. Lippincott,
l9h2. -

Anderson, H. H., and.H. F. Brandt. Study of motivation involving
self announced goals of fifth grade children and the concept of
level of aspiration. J. Soc. Psychol., 1939, 10, 209-232.

 

Bayton, J. A. Interrelationship between level of aspiration and
estimates of past performance. J, E52, Psychol., l9u3, 33, 1-21.

Bowman, Eugenia W. Level 9£_Aspiration and Frustratign_ig_5chizo-
phrenic and Normal Subjects. Unpublished Master's Thesis, Catholic
Univ. of America, 1959.

 

 

Cameron, N. The Psychology g§.Behavior Disorders. New York:
Boughton Mifflin, l9 7.

Chapnan, D. W. and J. Volkman. A social determinent of the level
of aspiration. .1. Abn. Soc. Psychol., 1939, 31+, 225-238.

Dembo, Tamara. Der Arger als Dynamisches Problem. Psych. Forsch.,
1931, 15, 1-1uh.

Eglash, A. Abnormal_Fixations. Unpublished Doctoral Dissertation.
Univ. of M1ch., 1951.

Escalona, Sibylle K. The effect of success and failure upon the
level of aspiration and behavior in manic-depressive psychoses.
Studies ig'Child Welfare, Iowa Univ., 1939, 17, 199-302.

Eysenck, H. J. and H. T. Himmelweit. An experimental study of the
reaction of neurotics to the experience of success and failure.
._J_. o_f_ Gen. Psychol., 19%, 35, 59-75.

Fenichel, O. The Psychoanalytic Theory of Neuroses. New York:
Norton, 19h5.

Festinger, L. A theoretical interpretation of shifts in level of
aspiration. Psychol. Rev., 19h2, 49, 235-250.

1H.

15.

16.

17c

18.

19.

21.

22.

23.

2h.

25.

26.

27.

23-

290

90

Frank, J. D. Individual differences in certain aspects of level
of aspiration. Amer. J, Psychol., 1935, #7, 119-128.

. The influence of the level of performance in one task
on the level of aspiration in another. J, Exp. Psychol., 1935,
18, 159-171.

. Some psychological determinents of level of aspira-
tion. Amer. J, Psychol., 1935, 47, 285-293.

' . A comparison between certain properties of level of
aspiration and random guessing. J,'Psychol., 1936, 3, 93-62.

. Recent studies of level of aspiration. Psychol.
131111., 19M, 38, 18-26.

_' . Level of aspiration test. In Murray, H. A. 23.9J,
Explorations J3_Personality, New York: 0xford.Univ. Press, 1938.

 

 

French, T. M. And analysis of the goal concept based upon study
of reactions to frustration. Psychoanalyt. 331,, 1991, 28, 61-71.

Freud, S. Neurosis and psychosis. In Freud, S. Collected Papers,
(Vol. 2). London: Hogarth Press, 1999.

Fromm-Reichman, Frieda. Transference problems in schizophrenics.
Psychoanalyt. Quart., 1939, 8, Ml2-417.

Gardner, J. W. Level of Aspiration in Response to'a prearranged
sequence of scores. J, Exp. Psychol., 1939, 25, 601-621.

__, The relatiOn of certain personality variables to
level of aspirations. J, Psychol.,.l9NO, 9, 191-206.

Gilinsky, Alberta S. Relative self estimate and the level of as-
piration. J, Exp. Psychol., l9u9, 39, 25b—259.

Gould, Rosalind. An experimental analysis of "level of aspiration."
Genet. Psychol. Monogr., 1939, 21, 1-116.

. Some sociological determinants of goal strivings.
J, Soc. Psychol., 1941, 13, 461-973.

 

. and N. Kaplan. The relation of level of aspira-

tion to academic and personality factors. J, Soc. Psychol., 19N0,
11, 31-ho.

Bruen, E. W. Level of aspiration in relation to personality fac-
tors in adolescents. Child Develpm., 19u5, 16, 181-188.

30.

31.

32.

33-

31k

37.

38.

39.
1‘00

1+1.

“‘2.

1&3,
m4.

“‘5.

91

Guilford, J. P. Fundamental Statistics in Psychology and Educa-
tion. New York: McGraw Hill, 1950.

Hausman, M. F. A test to evaluate some personality traits. J.
Gen. PsychoJ” 1933, 9, 179-187.

Heathers, Louise 3. Factors producing generality in the level of
aspiration. J. Exp. gummy, 19112, 30, 392-1106.

Hilgard, E. R., E. M. Sait and G. A. Magaret. Level of aspiration
as affected by relative standing in an experimental social group.

J. m. Psychol., 1940, 27, 1411-421.

Hilgard, E. R. and E. M. Sait. Estimates of past and future per-
formance as measures of aspiration. Amer. J. Psychol., 1941, 54,
102-108. A

Himmelweit, H. T. A comparative study of the level of aspiration
of normal and neurotic persons. Br. J. Psychol., 1947, 37, 141-59.

Hold, R. R. Motivational Factors in Levels if Aspiration. Unpub-
lished Doctoral Dissertation, Harvard U., 1943.

. Effects of ego-involvement on level of aspiration.
Psychiatry” 1945, 8, 299-317.

 

. Level of aspiration, ambition or defense. J. Exp.

Hoppe, F. Erfolg und Miserfolg. Psych. Forsch., 1930, 11+, 1-62.

Hunt, J. McV. and C. N. Cofer. Psychological Deficit. In Hunt,
J. Nov. (ed.) Personality and the Behavior Disorders. New York:

Hurlock, Elizabeth B. An evaluation of certain incentives used in
SChDOI'Orke £0 Educ. PSEChOIO. 1925, 16. 1%.1590

Irwin, E. W. and M. G. Mintger. Effects of different instructions
and motivation upon increases of level of aspiration. Amer. J.
Psychol., 19142, 55, lboo-1106.

Jenkins, R. L. The nature of the schizophrenic process. Arch.
Neurol. Psychiat., 1950, 63+, 2143-262.

Jones, I. Y. Tests of Hypotheses: One-Sided vs. Two-Sided Altern-
atives, Psychol. Bull., 1952, 1+9, 143-147.

Jucknat, M. Leistung, Anschpruchniveau und Selbstbewusstein.
Psych. Forsch., 1937, 14, 1-62.

46.

47.

49.'

50.

51.

52.

54-

55.
56.

57-

58.

60.

.92
Katz, I. Emotional expression in failure. J, Abn. Soc. Psychol.,
1950. 45. 329-350.

Kendig, I. Studies in perseveration. IV: Selective perseveration
and repetition. J, Ps chol., 1937, 3, 253-259.

Klugman, S. F. Emotional stability and level of aspiration. J,
Gan. PsyChOIQ. 19148. 38, 1013.118.

Kyle, G. T. J_Comparison 9: Normal and Schizophrenic Subjects in
Level 9: Aspiration. Unpublished Doctoral Dissertation, New York
University., 1948.

 

 

Lantz, Beatrice. Some dynamic aspects of success and failure.
Psychol. Monogr., 1945, 59, 1-39.

Lewin, K. gt_§J, Level of aspiration,in Hunt, J. McV., Personality
and the Behavior Disorders. New York: Ronald Press, 1944.

 

MacIntosh, A. Differential effect of the status of competing groups
upen level of aspiration. Amer. J, Psychol., 1942, 55, 546-555.

McGehee, W. Judgment and the level of aspiration. J, Gen. Psychol.,
1940, 22.,3_15,

 

McKinney, F., G. B. Strother, R. R. Hintes and R. A. Allee. Exper-
imental frustration in a group test situation. J, Abn. Soc.

Psychol., 1951, 46, 316-323.
Maier, N. R. r. Frustration. New York: McGraw Hill, 1949.

Marquart, Dorothy I. The pattern of punishment and its relation
to abnormal fixation in adult human subjects. _J, Jan, Psychol.,
1948, 39, 107-144.

Miller, D. R. Responses of psychiatric atients to threat of
failure. J. m. Soc. Psychol., 1951, _, 378-387.

Noyes, A. P. Modern Clinical Psychiatry. Phila.: W. B. Saunders
.Co., 1949.

 

O'Connor, J. Administration and Norms for the Finger Dexterity,
Test, Worksample l6, and the Tweezer Dexterity Test, Worksample
J2, Human Engineering Laboratory, Technical Report, No. 16.
Boston, Mass.: The Laboratory, 1938.

 

Preston, M. G. and J. A. Bayton. Differential effect of a socio-
logical variable on three levels of aspiration. J, Exp. Psychol.,

1941. 29. 351-369.

 

bl.

62.

63.

64.

b5.

66.

67.

68.

b9.

93

Rabin, A. I. A short form of the Wechsler-Bellvue test. J, App .
Psychol., 1943, 27, 320-324.

Rotter, J. B. Level of aspiration as a method of studying person-
ality. II: Development and evaluation of a controlled method.
J. Exp. Psychol., 1942, 31, 410-422.

 

. Level of aspiration as a method of studying person-
ality. IV: The analysis of patterns of response. J, Soc.

Psychol., 1945, 21, 159-177.

Sears, Pauline S. Levels of aspiration in academically successful
and unsuccessful children. J, Abn. Soc. Psychol., 1940, 35, 498-535.

 

Sears, R. R. Success and failure,in McNamar, Q. and Merrill, M.A.
(eds.) Studies in Personality, New York: IcGraw Hill, 1942.

Schaefer, R. The Clinical Application g£_P§ychologica1 Tests. New
York: International Universities Press, 1948.

 

Steisel, I. M. and B. D. Cohen. The effects of two degrees of

failure on level of aspiration and performance. J, Abn. Soc.
Ps chol., 1951, 46, 79.82.

U. 8. Veterans Administration. Nomenclature of psychiatric dis-
orders and reactions. T.B. 10A-78, 1947.

Webb, W. W. Conceptual ability of schizophrenics as a function of
threat of failure. (abstract), Amer. Psychologist, 1952, 7, 335.

|||||

 

w...
Mn...
0.
E
S
U
M
0
no
R
W

Ir :4