'

 

 

 

.o
o
u .
lo
0.
>0
Y
o

n
o
' \

-‘V

 

‘0“3‘

 

.o.w

u-“-.

- “‘—

g-

-o

-v-

 

 

 

ESSION AND PERCEIVED REINFORCEMENT: , . . .
ammsmn ' 1 .
mastortheoegreeom.p. -
MICHIGAN STATE unwmsm ‘
PAULA c. mm
1976 ‘ '

293

‘ a L‘! mu; m M! H: mm 1w 1% Lug) "’1 ”fl 1;: !fl u M...  

 

 

r)
- "13‘
31-21 3
IL .1,
1‘ '.

\;}
ABSTRACT

DEPRESSION AND PERCEIVED REINFROCEMENT: AN EXTENSION

By
Paula C. McNitt

The aim of this study was to further elucidate whether and
how depressed persons differ from nondepressed in their preception
of response contingent reinforcement. The study's main hypothesis
is that depressed individuals require a higher rate of reinforcement
in order to perceive these rewards as, in fact, contingent upon their
own actions.

The perception of reinforcement, because it is a private
event, can be assessed only indirectly. One means of doing so is to
measure trial by trial changes in verbalized expectancies of future
success following success and failures on an experimental task.

Rotter and his colleagues, (Phares, l957; Rotter, Crowne and Liverant,
1961) have found that people modify their verbalized expectancies more
on skill tasks, where rewards are response dependent, than on chance
tasks, where rewards are response independent. Thus one way of
determining whether and how depressed persons differ from nondepressed
in perceiving reinforcements is to compare their changes in verbalized

expectancies for success on skill and on chance tasks.

Paula C. McNitt

In this Depression X Tasks X Reinforcement, with repeated
measures on the Tasks factor, design study, changes in verbalized
expectancies on skill and chance tasks, at 50% and 75% reinforcement
levels were compared for depressed and nondepressed subjects. Three
measures of change were employed: change in expectancy following
the first, successful trial, verbalized expectancy following the
final, successful trial, and total amount of appropriate expectancy
change. Seven specific hypotheses were tested. These predicted a
task x depression x reinforcement interaction. Nondepressed subjects
would produce larger changes in expectancy on the skill task than
would the depressed subjects, but only in the 50% reinforcement con-
dition. When reinforced on 75% of the skill task trials, however,
the depressed subjects would respond no differently than the non-
depressed. On the chance task, there would be no difference between
depressed and nondepressed groups, at either reinforcement level.

The 24 women and 16 men who volunteered for this study were
recruited from summer session classes at Michigan State University.
The 20 nondepressed subjects had all scored 5 or below on the Beck
Depression Inventory: the 20 depressed subjects had all scored 9
or above on this scale. After determining each subject's level of
depression, the experimenter assigned him, at random, to one of four
Reinforcement X Task Order cells.

Each participant was allowed eight trials on both the skill
and chance tasks. The skill task consisted of manipulating the

Vertical Aspiration Board, a vertical frame in which a platform

Paula C. McNitt

block is set. The subject's task, here, is to raise the block to a
specified point on the frame without dislodging a steel bearing
placed on the platform. The chance task was a guessing game in which
the subject had to anticipate which of two marked slides would be
presented. The experimenter controlled the trial by trial outcomes
on each task. All subjects were permitted to succeed on the first
and last trials. Prior to each trial,each subject estimated, on a
ten point scale, how certain he felt that he would succeed on that
trial.

The results of this experiment were opposite to those pre-
dicted. On the skill task, the depressed subjects in no way differed
from the nondepressed in modifying their expectancies for success.
This was the case for both the 50% and 75% reinforcement groups.
Obviously, this finding for the 75% reinforcement group lacks the
import it would have had, had there been a difference between
depressed and nondepressed subjects in the 50% reinforcement group.
Thus, these findings do not support the hypothesis that depressed
persons differ from nondepressed in their perception of response
contingent reinforcement.‘ This study yielded an additional, and
wholly unexpected finding. The depressed subjects differed from the
nondepressed in modifying expectancies on the chance task. When
reinforced on 75% of the chance task trials, the depressed subjects
produced significantly higher final expectancies and larger total
amounts of expectancy change than did their nondepressed counter—

parts. This suggests that depressed persons differ from nondepressed

Paula C. McNitt

in their perception of response independent reinforcements. These
data suggest that depressed persons respond to chance task rewards
as if these were a product of their own skill.

The results of this study when considered together with
those of other studies (Miller and Seligman, l973) which have demon-
strated differences between depressed and nondepressed subjects in
their perception of response contingent reinforcement suggest a
broader conceptualization of the relation between depression and the
perception of reinforcement. This is that the depressed person
harbors a faulty notion of the roles played both by his own skill
and by factors outside of his control, and may misperceive either
response dependent or response independent rewards, or both. He may
either fail to recognize the rewards of his efforts as such or blame
or credit himself for rewards and punishments he does not, in fact,
control. This new conceptualization remains, of course, purely

hypothetical.

DEPRESSION AND PERCEIVED REINFORCEMENT: AN EXTENSION

By

‘5.»
n“-

\-.
Paula C. McNitt

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS
Department of Psychology

T976

To my husband, Andrew

11

ACKNOWLEDGEMENTS

I should like to extend my sincere thanks to my committee
members: Professors Dozier Thornton, Robert Zucker, and Bertram
Karon for their help and support in the preparation of this thesis.
Special thanks are due my chairman, Dozier Thornton. From the
planning stage through to the final draft of this thesis, his careful.
thoughtful critiques helped keep my thinking and writing clear. I '
should also like to thank Dr. Baron Perlman, now of the University
of Wisconsin, Oshkosh, for serving on the thesis committee during

the 1975-76 academic year.

TABLE OF CONTENTS

LIST OF TABLES
Chapter
I. INTRODUCTION .

A Description of Depression
Emotional Manifestations
Cognitive Manifestations
Motivational ManifestatiOns
Physical Manifestations .

Theories of Depression .

Operant- Behavioral .
The Learned Helplessness Analogue
The Cognitive Theory.

Depression and the Perception of Reinforcement .

II. METHODS
Subjects
Instruments
Apparatus .
Chance Task .
Skill Task
Procedure
Dependent Measures
III. RESULTS
IV. DISCUSSION
APPENDIX

REFERENCES .

iv

Page

Tables
1.

LIST OF TABLES

Beck Depression Inventory Scores, Age and Sex Char-
acteristics of the Two Experimental Groups

Means for Each of the Four Groups on Skill and
Chance Tasks . . . . . . . .

Analysis of Variance of Change in Expectancy from
Trial l to Trial 2 . . . . . . . .

Analysis of Variance of Expectancy Following the
Final Trial . . . . . . . . .

Analysis of Variance of Total Amount of Expectancy
Change . . . . . . . . . .

Analysis of Variance of Trial by Trial Changes in
Expectancy on the Skill Task - 75% Rf . .

Page

37

46

47

48

49

51

CHAPTER I

INTRODUCTION

This study investigates rate of reinforcement as a potential
mediator of the relation between depression and the perception of
reinforcement. The study's main hypothesis is that depressed indi-
viduals require a higher rate of response contingent positive rein-
forcement, than do nondepressed, in order to perceive these rewards
as response contingent. Two basic ideas underlie this hypothesis.
The first is that depression is intimately related to the availability
of positive reinforcers. A person becomes depressed if he loses
either an important source of or means to these rewards. The second
idea is that the reward value of any event, which is its capacity to
uplift mood and strengthen behavior, depends, in part, upon the indi-
vidual's recognition of this event as contingent on his behavior.
Thus, the depressed person is less likely to feel rewarded and
encouraged by the favorable outcome of his efforts because he is less
likely to be convinced of the causal relation between these two.

This hypothesis attempts a synthesis of three theories of
depression: operant behavioral, learned helplessness, and cognitive.
These theories, which shall be presented following a discussion of
the manifestations of depression, hardly exhaust the range of current

conceptualizations of the nature and etiology of depression. Other

I

psychological theories, pre-eminently the psychoanalytic, offer inter-
esting and useful approaches to the understanding and treatment of
depressive disorders. Medical theories, which focus on the physical
and physiological symptoms of depression, attend to the genetic and
biochemical bases of the disorder. The behavioral-operant, learned
helplessness, and cognitive theories are discussed because they
directly address the issue of the inter-relations between behavior,
environmental feedback, and depression. No one of the three theories
is offered as the correct one; the complexity of the phenomena argues
against a simple, all inclusive explanation. Rather, these theories
are presented as starting points from which to begin to make sense

of the phenomena of depression.

A discussion of "being depressed" is not unlike a discussion
of "having a virus." Hardly any of us has never had a cold, though
few of us have had viral pneumonia. Some of us are subject to more
frequent and severe colds than others. Viral symptoms "blossom" in
any number of combinations and intensities. Analagously, most of us
have felt some aspect of depression at some time in our lives, but
few of us have known severe, clinical depression. Some of us are
more prone to depressive episodes than others. So too, the manifes-
tations of depression appear in a wide range of combinations and
intensities. It is precisely this variability which makes it well
nigh impossible to formulate a concise definition of depression. It
has been referred to as a "mood state" a "symptom complex," and even

"a well defined clinical entity" (Beck, 1972, p. 6). The Diagnositc

and Statistical Manual II of the American Psychiatric Association
includes five separate categories of depressive disorder: manic-
depressive reaction, psychotic depressive reaction, schizo-affective
reaction, psychoneurotic depressive reaction, and involutional psy-
chotic reaction. These categories imply qualitative differences
among depressive disorders. Some clinical researchers have drawn a
dichotomous, qualitative distinction between neurotic and psychotic
depressions, and between process/endogenous and reactive depressions.
Others have suggested that the neurotic-psychotic and process/
endogenous-reactive distinctions are differences of degree rather
than of kind. Lewis, Mapother, and Hoch represent the latter camp;
Gillespie, and Carney and Garside represent the former. [See Beck,
1972, pp. 64-67 and Becker, 1974, pp. 11-13, 41-48 for discussions

of the issues and debates on the classification of depression.) Thus,
one may question whether it is legitimate to refer to "depression" in
general.

The very fact that these several disorders are all called
instances of depression suggests that there is some continuity and
similarity among their manifestations. Rather than wrangle with the
ontological and etiological implications of a classificatory system
for depression, this paper shall attempt only to describe depressive
disorders. Following Beck's example, the manifestations of depres-
sion shall be described without reference to any underlying structure
or cause. This paper will construe depression as a psychopathologi-

cal dimension consisting of a number of behavioral and affective

manifestations. These symptoms appear in a variety of combinations
and in a wide range of intensities. No one symptom is necessary for
a diagnosis of depression. Any individual case of depression is

best defined in terms of its position on the dimension, that is, in
terms of the number and intensity of depressive symptoms. Although
this investigation makes no attempt to resolve the issue of the unity
versus plurality of depressive disorders, it shall restrict itself to
those instances of depression of neurotic, as opposed to psychotic,

intensity, for which enVironmental causes are presupposed.

A Description of Depression

 

The manifestations of depression are, to speak metaphorically,
negations of basic life instincts towards activity and gratification.
The experience of depression involves sadness, pessimism, apathy, a
reduction in constructive activity, guilt feelings, self-devaluation,
and loss of appetite and sexual desire. In a sense, depression con-
stitutes a major breakdown in the functional interaction between the
individual's behavior and environmental reinforcement. The depressed
person drastically constricts his range of productive activities,
believes he can do no better for himself, and feels his efforts are
unrewarded, if not actually punished. Thus, depression estranges
the individual from his customary life style of activity and inter-
personal involvement. The paradox of depression is that it may be
quite unwarranted by the person's actual life circumstances. Although

dePressjon has traditionally been understood as a primary mood

disorder, its manifestations are best categorized as emotional, cogni-

tive, motivational, and physical.

Emotional Manifestations

 

The emotional signs of depression are the most obvious and
familiar. Typically these include feelings of sadness and loneli-
ness. Their pervasiveness and severity vary considerably. In cases
of mild depression, one's mood may fluctuate during the day. A joke
or compliment may relieve the "blueness." In more severe cases,
however, the depressed patient feels overwhelmed by his mood; he
simply cannot snap out of it. Crying spells become more frequent
during depressive periods, especially among women. The inability to
derive any gratification from one's efforts is a hallmark of depres-
sion. When a depression deepens, that is, when the feelings of
sadness, helplessness and apathy intensify, the range of activities
which afford the individual pleasure also contracts. For example,
the depressed patient typically loses his capacity for humor. While
he may perceive the point of a joke, he may not be able to respond
to it. Emotional attachments may also be neglected or relinquished
during depressive episodes. Depressed patients commonly report that

they can neither love nor care for their friends and relatives.

Cognitive Manifestations
The cognitive manifestations of depression indicate that the
depressed person systematically misinterprets his experiences. The

degree of distortion varies with the severity of the depression. In

the most severe cases, such misinterpretations can assume the rigidity
and intensity of delusions. The depressed person maintains a negative
view of himself, his world, and the future. He sees himself as defi-
cient and deprived. Ironically, he especially devalues those attri-
butes which have traditionally afforded him pleasure and rewards. A
beautiful woman will judge herself ugly; a talented academician
decides he is really very stupid. The depressed person feels ineffec-
tive. Experimenters who have worked with severely depressed patients
often report that such a patient's initial response to a task is that
he cannot possibly do it (Seligman, 1974). The depressed person
magnifies the significance of any and all incidents. Minor setbacks
are catastrophes: errors are sure signs of personal inadequacy. Not
only does the depressed person feel inadequate, but he also blames

and criticizes himself for this. In milder depressions, this amounts
to intolerance of one's shortcomings. In more severe cases, this
self-blame extends to things for which the person is in no way respon-
sible. As an illustration of this, Beck (1972, p. 232) relates the
case of a woman who, when she took her children on a picnic and it
stormed, blamed herself for not choosing a better day. The depressed
person usually feels deprived of the love and material possessions

he needs. Depression generally involves pessimism: the feeling that
one's current failures and unhappiness are permanent and irreversible.
As Melges and Bowlby (1969) point out, this negative attitude towards
the future, in turn, influences and structures the person's thinking
and behavior. In short, the attitudes of helplessness and hopeless-

ness are central to depression.

Motivational Manifestations

 

The motivational manifestations of depression are regressive
in character. The depressed person prefers passivity and dependency,
and chooses those activities which demand little energy or responsi-
bility. The decrement in spontaneous, productive activity is another
hallmark of depression. In severe cases this inertia may spread to
the most simple and commonplace activities, such as getting up out of
bed in the morning and getting dressed. A severely depressed person
nay know very well what he ought to do, but feel incapable of doing
so. Beck (1972, p. 28) has labelled this as a "complete paralysis of
the will." The wish to escape difficulties and to avoid responsi-
bility is commonplace in depression. In mild cases of depression,
one procrastinates and seeks relief in such passive diversions as
watching television or drinking. Escapist tendencies assume the most
extreme form in suicidal wishes. Depression entails the desire not
just to avoid responsibility, but also to receive help. Melges and
Bowlby (1960) suggest that the depressed person feels he must rely
on others because he believes himself unable to solve his problems
on his own. Beck comments, though, (1972, p. 32) that this desire
for help is more than a reaction to helplessness, and is, rather,

experienced as a positive craving for help.

Physical Manifestations

 

The physical manifestations of depression are quite concrete

and quantifiable; however, these correlate weakly with each other and

with clinical ratings of depression. Beck (1972, p. 33) reports
correlation coefficients ranging from .20 to .35 between ratings of
depth of depression and such physical symptoms as fatigability,
sleep loss, and loss of appetite. The loss of appetite, often a
first sign of depression, may develop into a veritable aversion to
food. Food is not the only thing which loses its appeal; depressed
patients commonly report a loss of interest in sex. Sleep disturb-
ance is another common symptom of depression. There is strong evi-
dence (see Beck, 1972, p. 34) derived from both direct observation
and EEG recordings, that depressed patients sleep less and experience
more restlessness while asleep than do other medical patients.
Depressed persons often report that they waken earlier from their
less restful sleep. Curiously, however, sleep disturbance is mini-
mally related to the increased fatigability of which many depressed
patients complain. Beck reports (1972, p. 36) a correlation of only
0.28 between reported sleep disturbance and fatigability. There
appears also to be diurnal variation in this feeling of tiredness;

it is most pronounced at waking and diminishes as the day wears on.

The most noticeable physical manifestation of depression is a
change in activity level. Depression may be either retarded or agi-
tated. Retarded depression entails a gross reduction in spontaneous
activity. Movements and gestures are minimal and slow. In the moSt
extreme cases, this retardation is a virtual stupor. Verbal output
is reduced; it may degenerate to grunts and groans. Agitated depres-

sion, on the other hand, is characterized by ceaseless, but

purposeless, activity such as pacing, handwringing, and constant
talking or mumbling. The emotional tone of the agitated person is
one of anguish and frenzy. He is incapable of channeling his energy

into constructive behavior.

Theories of Depression

 

Operant-Behavioral

 

The operant-behavioral theory of depression eschews the notion
that depression is an illness and assumes that it can be described
and explained by reference to the same concepts and behavioral prin-
ciples as any other learned psychological disorder. Depression is
best described quantitatively; that is, in terms of the relative
frequencies of various behaviors. Thus, depression constitutes a
continuum of intensities of a psychological state. Normal low mood
and clinical depression are two points on the same psychological
dimension. Secondly, depression is explicable in terms of the envi-
ronmental stimuli which elicit and maintain it.

A topographical description does not distinguish the behavior
of a depressed person from that of anyone else (Ferster, 1974, p. 34).
There are no uniquely depressive behaviors. Rather, it is the rela-
tive frequency of various behaviors which differentiates among indi-
viduals. Depression is characterized by a drastic reduction of
certain behaviors, particularly of active, productive behaviors. As
noted earlier, the depressed person avoids demanding work and with-

draws from social interaction. Lewisohn and his colleagues (see

10

Lewisohn, 1974) have stressed the presumably causal relation between
depression and deficient and distorted social interaction. One of
their studies (Libet and Lewisohn, 1973) found that in a small group
setting, depressed subjects emitted interpersonal behaviors at a rate
only one half that of nondepressed Subjects. Other pleasurable pur-
suits also decrease, if not actually drop out of the depressive's
response repertoire. Depressive episodes are also marked by an
increase in unpleasant, and seemingly purposeless behaviors. Examples
of these are statements of dysphoria, self-depreciation and guilt,
somatic complaints, suicidal wishes, and, in some cases, agitated
responses such as constant mumbling, pacing, and handwringing.
Changes in environmental reinforcement contingencies can
account for these changes in behavior. Insofar as he is emitting
fewer constructive behaviors than previously, the depressed person
is presumed to be on a massive extinction schedule. The environment
is simply no longer adequately rewarding his customary behaviors. As
Ferster cautions (1974, p. 35), however, to say this is not equivalent
to specifying the cause of depression. A number of factors can
account for a reduced rate of behavior. First, the type of schedule
on which a behavior is reinforced influences it considerably. Sched-
ules which require a large, fixed amount of work prior to reinforce-
ment ultimately weaken behavior, and prompt long periods of inactivé.
ity. Variable schedules, though, are less likely to strain the
individual (Ferster, 1974, p. 36). A shift to a fixed reinforcement

schedule may, in part, explain the work-related depression many

11

people experience in middle age. At this time, one has probably
achieved a plateau in his profession, a stable position of responsi-
bility which offers little variation in work requirements. One must
settle into doing a good deal of work for unchanging rewards. Such

a pattern of reinforcement generally weakens behavior, and may even
prompt feelings of dysphoria. Radical changes in the environment
also alter the reinforcement patterns which have shaped and maintained
the individual's response repertoire. Death or separation from a
loved one, a financial or professional setback, or even ordinary
retirement from work may constitute a loss of a significant source of
reinforcement. The more one has relied on a single person or job to
provide rewards, the more profound will be the effects of that loss.
Thus, the depressed person's reduced spontaneous productive activity
and social interaction is a natural, predictable consequence of his
interaction with a relatively unrewarding environment.

The operant theorists' explanations of the increase in depres-
sive behaviors such as low-mood, self-depreciation, delusional think-
ing, and agitated pacing and handwringing, however, refer to
additional assumptions. A behavioral theory assumes (Lewisohn, 1974,
p. 158) that a low rate of response contingent positive reinforcement
is itself an eliciting stimulus for feelings of sadness, fatigue,
and apathy. In this sense, depressive behaviors are respondents.

The cognitive manifestations of depression, such as low self-esteem
and pessimism, are secondary elaborations of one's feelings of sadness.

(Lewisohn, 1974, p. 160). These are attempts at explanations of one's

12

emotional state. For example, a depressed person can explain his
sadness as a natural consequence othis unworthiness. His blue mood
must be a punishment for his wrongdoings. The bizarre, agitated,
apparently purposeless behaviors associated with depression can be
explained as emerging from an impoverished response repertoire.
Ferster (1974, p. 32) has suggested that their emergence results
simply from the lack of significant competing activities. Bizarre,
purposeless behaviors are apt to appear in any situation where normal
activity is restricted. A depressed patient may resort to pacing,
picking at himself, and constant mumbling simply because his other
constructive behaviors have been extinguished. Thus, these agitated
behaviors are neither strongly motivated nor strongly maintained. A
significant ramification of this outlook is that depressive-agitated
behaviors, while diagnostically relevant, are not necessarily signs
of a more serious underlying pathology. .

Once elicited, depressive behaviors are further shaped and
maintained by their environmental consequences. Complaints, crying,
sadness, and protestations of worthlessness and guilt are reinforced
by the attention and solicitude they elicit from the depressive's
family and friends. Liberman and Raskin (1971) provide clinical evi-
dence which supports this. These researchers report, as an example,
the case of a depressed housewife whose family members were asked to
record the woman's sadness, complaints, and crying, attending to
these as usual, then to ignore these, and finally, to resume baseline

observations. As anticipated, the frequency of these behaviors

l3

dropped considerably while being ignored and returned to baseline
while being attended to. Burgess (1968) similarly reports successful
extinction of depressive behaviors through the therapist's inten-'
tional inattention to these.

The parsimony of a behavioral theory of depression, which
makes the model so attractive, is also its major shortcoming.
Because the behavioral model refuses to account for the inner workings
of the "black box," two of its tenets remain unsubstantiated. These
are: the loss of response contingent reinforcement is an eliciting
stimulus for depressive affect, and, purposeless, agitated behaviors
are mere "fillers" of an otherwise diminished response repertoire.
Neither of these assumptions is intuitively obvious: why should the
loss of reinforcement evoke sadness, rather than anger, for example?
In the same way, the behavioral model fails to account for the moti-
vational manifestations of depression: why does the depressed person
actively avoid and escape difficult and demanding work? Can changes
in reinforcement contingencies of themselves account for the paralysis
of the will, which is a hallmark of depression? Even more puzzling
from the behavioral viewpoint is the depressed person's inability to
derive gratification from those activities and social relations
which he had formerly found satisfying. Why does the depressed per-
son feel he cannot love or care for others once dear to him? In
short, the behavioral-operant model, by its very nature, cannot ade-

quately deal with depression as a lived experience.

14

The Learned Helplessness Analogue

 

An operant theory holds that depression follows from the loss
of response contingent reinforcement. Seligman's learned helplessness
theory (Seligman, 1974) argues that depression results not simply
from the loss of reinforcement, but rather from the loss of control
of both positive and negative reinforcers. Although Seligman's the-
ory is more specific, it is also the more modest of the two. First,
it attempts to account for only one type of depression; namely,
reactive depression which is characterized by passivity, feelings of
helplessness and powerlessness, and negative expectations for the
future. Secondly, insofar as learned helplessness in animals is an
experimentally induced phenomenon it can serve, at best, as an ana-
logue for depression in man. Learned helplessness and depression,
though, can be compared in terms of their behavioral manifestations,
etiology, cure, and prevention.

Defined behaviorally, learned helplessness is the interfer-
ence with adaptive responding which results from experience with an
uncontrollable aversive stimulus. More specifically, this means
passivity in the face of trauma and retardation in subsequent learn-
ing of responses that can, in fact, produce relief. Overmier and
Seligman (1967) and Seligman and Maier (1967) found that dogs given
inescapable/unavoidable shock while restrained in a Pavlovian harness
were later deficient in escaping shock when allowed to do so (in a
shuttle box). A naive dog subjected to shock in a shuttle box will

respond initially by howling, defecating, and running about, but

15

eventually will accidentally jump the barrier and thereby escape the
shock. After several such trials, the dog will have learned to
immediately jump the barrier to escape the punishing stimulus. Dogs
which are first given inescapable shock do not readily learn to
‘escape/avoid. When placed in the shuttle box these animals may at
first react to shock as do naive dogs, but eventually they simply lie
down and whine, passively accepting their punishment. If a dog does,
by chance, jump to escape, it generally fails to learn from this
experience. Rather than repeat the response and thus learn that it
produces relief, the dog reverts to its passive stance.

Substantial evidence argues for the generality of the help-
lessness phenomenon. Seligman's (1972) review of the literature
notes that only 6% of experimentally naive dogs, but two thirds of
150 dogs first subjected to inescapable shock, failed to learn to
escape shock in a shuttle box. The helplessness phenomenon is not
restricted to dogs; it has been demonstrated in rats, cats, fish,
mice, and men (Seligman, 1974, p. 87). Experience with inescapable
shock interferes with other adaptive responses such as food getting
behaviors, pain elicited aggression, as well as a variety of avoid-
ance behaviors (Seligman, 1974, p. 87). That learned helplessness
has also been demonstrated in human subjects lends credence to it as
an analogue for human depression. Thornton and Jacobs (1971) found
that subjects who were first given reaction time task training to
escape shock were significantly faster on a second shock avoidance

task than were their yoked controls who had received equivalent

16

amounts of inescapable Shock. 0n post experimental questioning 60%
of the subjects given inescapable shock said they felt they could
not control the shook and therefore did not try to do so on test
trials. Of the subjects given initial escape training, though, 70%
reported that they thought they could control the shocks. Similarly,
Hiroto (1974), employing loud noise as the aversive stimulus, found
that his subjects who first suffered inescapable noise were signifi-
cantly poorer in learning a response to terminate noise than were
subjects first given either escape training or no noise at all.

The behaviors which define learned helplessness parallel

several major symptoms of depression. "Helpless animals are passive

in the face of punishment. Their responses are retarded and reduced
in amplitude. Analagously, the depressed person's behavior is char-
acteristically passive and retarded. Secondly, "helpless" animals
are slow to learn that their responses can produce relief. If a
"helpless" dog does, by chance, escape shock by jumping the shuttle
box barrier, it is unlikely to repeat the response. In order to cure
chronically "helpless” dogs, Seligman, Maier, and Greer (1968) had to
literally drag-them from side to side in the shuttle box from 20 to 50
tines before the dogs began to respond on their own. Analagously,
depressed people appear to be convinced of the inefficacy of their
actions. Experimental data comparing depressed and nondepressed psy-
chiatric patients (Loeb, Beck and Diggory, 1971) indicates that the
former'are significantly more pessimistic in estimating the proba-

bility of their success on an experimental task. The depressed person

17

consistently anticipates defeat and misconstrues his performances
accordingly. Moreover, the individual suffering through a depres-
sion is convinced that his failures, frustrations and unhappiness
will continue indefinitely. He feels helpless and hopeless. In
short, he believes, and acts as if, he has no control over the
rewards and punishments he receives.

The etiologies of learned helplessness and of depression are,
to some extent, comparable. Because the former is an experimentally
induced phenomenon, it is possible to specify the conditions which
produce it. Learned helplessness results from experience with uncon-
trollable trauma. In any learning situation, an animal presumably
learns two conditional probabilities: the probability of reinforce-
ment given a certain response, p (RFT/R), and the probability of
reinforcement given the absence of that response, p(RFT/R). A
response controls a reinforcer if and only if p(RFT/R) is greater
than p(RFT/R). Reinforcement is response independent if these two
probabilities are equal. For the dog restrained in the harness,
shock termination is the relevant reinforcer. Seligman (1974, p. 95)
has argued that what the dog learns here is that for any possible
response, neither its emission nor omission will terminate shock.

The crucial variable, then, is not the shock per se, but
rather, its uncontrollability. Seligman and Maier (1967) found that
dogs which, though restrained, were able to press a panel to termin-
ate shock, subsequently learned to jump the shuttle box barrier.

Their yoked controls, which received an equivalent amount of

18

inescapable shock failed to learn this adaptive response. Alterna-
tive explanations of the helplessness phenomenon such as adaptation
to shock or learned passivity have been disqualified by experimental
data (Seligman, 1974, pp. 95-96). Maier's (1970) findings discredit
the hypothesis that the animals learn passive responses antagonistic
to barrier jumping when they are shocked while restrained. Maier
(1970) placed dogs in so confining a chamber that they could termin-
ate shock only by remaining perfectly still. When these dogs were
later shocked in the Shuttle box, they first reacted by becoming very
still, but then learned to jump the barrier. Thus, the dogs' passiv-
ity was not a Sign of helplessness, but rather, an adaptive response.
A larger implication of this is that inactivity per se need not lead
to depression, as an operant theorist might suggest. Intentional
passivity can be a highly adaptive, controlling behavior. The reason-
able conclusion, then, is that it is the experience of being unable
to control a punishing situation which causes helplessness.

The causes of depression in man are less readily specifiable
than those of helplessness in animals. The learned helplessness
theory offers interesting possibilities. Seligman (1974, p. 98) has
suggested that it is precisely the loss, and particularly the sudden
loss, of control of a significant aspect of one's environment which
precipitates a depression. While there is no human equivalent to
being shocked while restrained in a Pavlovian hammock, there are
many severe stresses which, apparently, precipitate depressive epi-'

sodes. Usually, these are sudden losses or failures, such as the

19

death of or separation from a parent or spouse, a financial or aca-
demic failure, or physical illness or injury (see Beck, 1974, p. 7).
The impact of such a stress is two-fold insofar as it entails the
loss of a major source of positive reinforcement, as well as emotional

pain. This suggests that depression follows from the loss of control

. its

not just of punishments, but also of rewards. That depression attends
the loss of control of positive reinforcements is as yet a matter of
speculation; the helplessness literature thus far has focussed on

control of aversive stimuli. Seligman (1974, p. 100) has proposed

 

that one can become depressed by receiving noncontingent rewards. He ;
suggests that this is the case with beautiful women and successful

executives who suffer depression. They are rewarded for who they are

rather than for what they do. In that sense, their actions no longer

control their environment's responses to them. One thorny problem

which remains for the learned helplessness theory is that loss of a

significant source of reinforcement does not necessarily lead to

depression. Not everyone reacts to loss or failure by becoming clini-

cally depressed. The question to be considered next is: what differ-

entiates those pe0ple who become depressed from those who do not?

The developmental/learning history of the depression-prone
individual may be a critical variable. Beck (1974, p. 7) discusses
two types of childhood experiences which predispose a person to
depression. They are the sudden loss of a loved one, and the adop-
tion of perfectionistic standards for personal achievement. Both of

these types of experiences teach the child that he is helpless in the

20

face of his environment. Traumatic loss, particularly the loss of a
parent, may sensitize the child to losses in general. Such a trauma
can predispose one to react to any loss or setback in the same extreme
way as he did when a child (Beck, 1972, p. 278). In this sense,
depression is akin to a conditioned response to loss or failure. In
a study comparing depressed and nondepressed psychiatric patients,
Beck and colleagues (reported by Beck, 1972, pp. 226-227) determined
that the incidence of childhood loss of a parent was twice as high
among depressed patients. The incidence for the depressed group was
27% and 12% for the nondepressed. Yet childhood loss obviously is
not a necessary condition for susceptibility to depression. Helpless-
ness and hopelessness can be learned without trauma. Beck notes
(1974, p. 7) that depression prone individuals are also those who
because they learned rigid, perfectionistic standards for themselves
during childhood, necessarily leave themselves vulnerable to feelings
of defeat and failure. Viewed from the learned helplessness model,
one is bound to see his action as ineffective if he harbors exces-‘
sively high standards of success. One will always feel helpless to
achieve one's goal, if one's goal is perfection. It appears, then,
that the depression prone individual is one who has learned, either
traumatically or insidiously, that he has little likelihood of
controlling his environment.

Both the operant and learned helplessness theories explain
depression as a reaction to radical changes in environmental rein-

forcement contingencies. These theories have been generated, for the

21

most part, from experimental research with animals. Thus it is
hardly surprising that these paradigms do not incorporate such
"distinctly human" variables as cognitions and values. Yet it is
quite reasonable to suppose that these variables do mediate the
effects of environmental forces in the development of depression.
Aaron Beck has developed a theory of depression which places cogni-

tion in a central causative role.

The Cognitive Theory

 

Depression has traditionally been treated as a primary mood
disorder (Beck, 1971). The cognitive distortions accompanying
depression have been relegated to the status of by-products of the
affective disorder. Beck (1971, 1972) has argued the opposite:
depressive affect follows depressive cognitions. His primary assump-
tion is that one's emotional reaction to any situation depends on
his appraisal of it, regardless of the veracity of that appraisal.
What distinguishes a normal from a pathological affective response is
precisely the degree of correspondence between the actual situation
and the conceptualization one forms of it. Faulty conceptualizations
provoke excessive and inappropriate emotional responses. Pathologi-
cal emotions are actually reasonable responses to erroneous appraisals
of one's experiences.

To determine which life experiences depressed persons mis-
construe, and how they misconstrue these, Beck (1972, pp. 229-238)
analyzed compared the thematic content of psychotherapy sessions of

50 depressed and 31 nondepressed psychiatric outpatients. Beck noted

22

that the depressives' dominant themes focussed on irretrievable loss,
failure, deprivation, and personal deficiency. The depressed patients
constantly berated and blamed themselves, complained of overwhelming,
interminable problems, and proclaimed their inability to handle these.
They downgraded their abilities and acquisitions, magnified their
failures and ignored all their successes. Beck concluded that the
depressive's thinking is dominated by the triad of negative views of
the self, the world, and the future.

The distortions implicit in depressive cognitions are not
haphazard; rather, they are systematic errors. The processes of
depressive thinking can be classified as paralogic, stylistic, or
semantic (Beck, 1972, p. 234). Paralogical processes include arbi-
trary inference, selective abstraction, and overgeneralization.
Arbitrary inference is drawing a conclusion in the absence of support-
ing evidence. An example of this is interpreting the hurt or angry
facial expression of a stranger as a sign of disapproval. This kind
of depressive thinking is egocentric because it fails to consider
alternative explanations. Selective abstraction involves focussing
on a detail out of context. Beck (1972, p. 234) provides the example
of one of his patients, who, when her supervisor requested that she
no longer make extra carbon copies of his letters, concluded that he
was generally dissatisfied with all of her work. Overgeneralization
refers to drawing a general conclusion on the basis of a single fact
or event. As an example of the depressive's tendency to overgeneral-

ize, Beck (1972, p. 235) cites the case of a patient who, within half

23

an hour, judged himself a poor father because his children misbe-
haved at breakfast, an inadequate driver because other cars passed
him on the road, and an undedicated worker because several of his
colleagues were already at their desks when he arrived at work.
Stylistically, depressive thinking is characterized by exaggeration.
The depressed person magnifies all problems into catastrophes and
reduces his talents and accomplishments to insignificance. The seman-
tic distortion implicit in depressive cognition is that of inexact
labelling: that is, calling some thing or event by an inappropriate
name. Each of these types of distortion involves a negative inter-
pretation of the situation. It is this negativity which engenders
the emotions associated with depression.

Depressive cognitions are the product of the interaction
between certain types of cognitive schema and certain kinds of expe-
rience. A cognitive schema is a structure by which one screens,
codes, and evaluates information. The interaction of a schema with
incoming information produces a cognition. This cognition can be an
idea, judgement, wish daydream, self-instruction. Schemas vary both
in their content and their congruency with reality. That is, schemas
vary with respect to the veractiy of the cognitions they generate.
Beck (1972, pp. 282-285) has speculated that the depression-prone
person has develOped highly idiosyncratic schemas which are activated
by experiences of loss and failure. The products of this activation
are those characteristically negative cognitions involving arbitrary

inference, selective abstraction, and overgeneralization. As a

24

depression develops these schema dominate more and more of the
person's thought processes. These schema are thus applied inappro-
priately, even to the point where virtually any stimulus input can
activate a depressive cognition (Beck, 1972, p. 286). In this way,
the depressed patient's thinking becomes entirely incongruent with
his actual experiences.

Beck's cognitive theory further contends that the emotional
and motivational manifestations of depression follow from depressive
thinking. Beck's argument for this rests on his own experience with
depressed patients. Beck (1972, p. 237) has asked his therapy
patients to work backwards from their feelings of sadness, guilt, and
humiliation and to try to focus on the precipitants of these feelings.
Usually, these clients report that an unpleasant thought preceeded
the emotion. Typically, this thought involved the kind of content
and error characteristic of depressive thinking. The motivational
manifestations of paralysis of the will, escapist tendencies, suici-
dal wishes, and increased dependency needs are also the products of
specific types of depressive cognition. The lack of spontaneous
motivation follows from the depressive's pessimism. Because he
anticipates only failure and deprivation, he remains unmotivated to
undertake any strenous activity. The depressed person wants to avoid
any responsible task, for he sees this as overwhelmingly difficult.
Suicidal wishes constitute the most extreme expression of the desire
to escape. The depressed person seeks not only to be relieved of

his problems, but also to be helped and taken care of. These

25

increased dependency needs can be attributed to the depressive's
feelings of inadequacy and pessimism. In short, the depression-prone
person's style in assessing himself and his predicament condemns him
to misery, passivity, dependency, and escapism.

An intellectual understanding of the mechanisms of depressive
thinking, unfortunately, does not immediately liberate one from
depression. There are several reasons why depreSsive cognitions are
especially difficult to deal with. First, depressive cognitions are
automatic and involuntary; they seem to arise without reflection
(Beck, 1972, p. 236). Beck's patients have reported that they often
are unable to ward off these thoughts, even though they had resolved
to do so (Beck, 1972, p. 236). This is especially the case with sui-
cidal thoughts. Then, too, the depressed person finds his own
thoughts quite plausible. He cannot see their falsity and irration-
ality. Indeed, once he can question their veracity, their affective

value is reduced (Beck, 1972, p. 237).

Depression and the Perception of Reinforcement

By any theoretical account, the depressed person experiences
his life and his world as singularly unrewarding. His assessment,
however, does not necessarily accord with the reality of his predica-
ment. This is because none of us lives in complete accordance with
"objective reality." Rather, the reality each of us experiences is
also shaped and colored by one's values, needs, attitudes, and con-
cepts. Events are neither inherently punishing nor inherently reward-

ing. Thus, any one person's behavior is not necessarily reinforced

26

by events that are customarily thought rewarding. Indeed, even so
orthodox an operant theorist as Ferster concedes that internal,
organismic variables modify the effects of environmental reinforce-
ment contingencies. Ferster (1974, p. 39) comments that the indi-
vidual's ability to be rewarded by his environment depends on his
perception of it. Unless one accurately perceives the consequences
of his (and others') actions, he is likely to persist in unrewarded
behaviors and to fail to adopt rewarded ones. Extending this consid-
eration to the topic of clinical depression, the question which
arises is whether, and to what extent, the depressed person differs
from the nondepressed in his perception of the rewards and punish-
ments he receives. Researchers espousing the learned helplessness
and cognitive theories have addressed this issue.

In his discussion of the cognitive set implicit in learned
helplessness, Seligman (1974, p. 95) emphasizes that inescapable
trauma teaches both animals and men that their responses do not
control reinforcement. The depressed person, then, views his rewards
and punishments as independent of his own actions. Miller and Selig-
man (l973) found that depressed students reacted to their response
dependent rewards as if these were response independent. In other
words, depressed people view their rewards and successes as if they
were the product of chance or good luck.

The perception of reinforcement, because it is a private
event, can be assessed only indirectly. One means of doing so is to

measure trial by trial changes in expectancy of future success

27

following successes and failures on an experimental task. Rotter and
his colleagues have found that subjects change their verbalized
expectancies differently on skill and chance tasks. Rewards on a
skill task are response dependent, and on a chance task, response
independent. Verbalized expectancies are usually expressed as a
numerical self-rating of degree of certainty of success. Generally,
subjects utilize feedback, that is, positive or negative reinforce-
ment, in adjusting expectancy levels to the extent that they perceive
these reinforcements as consequences of their efforts. Phares (1957)
found that subjects given "skill" instructions for an experimental
task produced significantly more and larger changes in expectancy
level than did subjects given "chance" instructions for the same
task. In a more comprehensive study, Rotter, Crowne, and Liverant
(1961) compared changes in expectancy on separate skill and

chance tasks under four reinforcement conditions: 25%, 50%, 75%

and 100%. In this study, [as well as those of Phares (1957) and
Miller and Seligman (1973)] "success" and "reinforcement" are con-
sidered equivalent terms. Although these experimenters also used
small monetary rewards, they considered knowledge of results to be the
primary reinforcing stimulus. Rotter et al. (1961) determined that,
while expectancies prior to the first trial were equivalent for

skill and chance groups, skill task subjects raised expectancies
following the first, successful trial significantly higher than did
chance task subjects. Verbalized expectancies following the final,

reinforced trial were significantly higher for skill than for chance

28

group subjects. Total change over trials was significantly greater
for skill than for chance subjects, in the 25% and 50% reinforcement
groups. The failure to demonstrate differences on this measure for
the 75% and 100% conditions was ascribed to the tendency for expec-
tancies to approach asymptote on the last, consistently reinforced
trials. Miller and Seligman (l973) reflecting on this data, reasoned
that differential expectancy changes on skill and chance tasks would
demonstrate that depressed persons differ from nondepressed in their
perception of reinforcement contingencies.

Miller and Seligman (l973) hypothesized that depressed sub-
jects generally perceive the reinforcements they receive as response
independent. While this perception is correct for a chance task, it

is erroneous for a skill task. Consequently, depressed subjects

would not modify their verbalized expectancies on a skill task accord-

ing to their successes and failures. Nondepressed subjects, however,
accurately perceive their rewards as response contingent in a skill
task and as response independent in a chance task. Miller and Selig-
man's (1973) principal hypothesis, then, was that nondepressed sub—
jects would modify expectancies more than would depressed subjects
on the skill task. On the chance task, these groups would not differ.
These researchers also predicted that depth of depression would be
inversely related to accuracy of perception of reinforcement on a
skill task.

Miller and Seligman's (1973) study compared two groups of

college student volunteers. Subjects were assigned to either the

’ We»?! .' (2:. Jr"!

 

29

depressed or nondepressed groups according to their scores on the
Beck Depression Inventory (BDI). Their cutting score was 9. Accord-
ing to Miller's earlier research (unpublished thesis referred to by
Miller and Seligman), 9 is the mean score for college students on
this scale. Although the mean 801 score of 12.5 for the depressed
groups falls well below scores associated with clinical depression,
Miller and Seligman (1973) demonstrated significant differences
between this group and the nondepressed group. Their experiment
required each subject to perform both a skill task and a chance
task. The chance task was basically a guessing game, consisting of
ten trials of five presentations of slides marked either "X" or
"O." The subject's task was to anticipate which slide would appear
next. The experimenter determined the outcome of each trial by
manipulating his presentation of "X"s and "O"s. The skill task
required the subject to manipulate the "vertical aspiration board."
This equipment consists of a tilted platform block set in a vertical
frame. The platform can be raised and lowered by a pulley string.
The task is to raise the platform near to the top of the frame with-
out dislodging a steel bearing resting on that block. This bearing
is actually held in place by a concealed electromagnet which the
experimenter can switch on and off. In this way, the experimenter
controlled each subject's performance. All subjects were allowed to
succeed on five of the ten trials. For all subjects, two of these
five success trials were the first and the last.

Prior to each trial, the subject rated his certainty of suc-

cess. The eleven point scale assigned "0'I to certainty of failure

3D

and "10" to certainty of success. The three dependent measures were
change in expectancy level following the first, reinforced trial,
final expectancy level following the tenth, successful trial, and

the sum of the absolute values of all shifts in level of expectancy.
These last two were interpreted as measure of overall change in
expectancy. As predicted, on the skill task nondepressed subjects
showed significantly greater changes than did depressed subjects,

but only on the first two measures. Nondepressed subjects increased
expectancies following success trials, but did not lower expectancies
following failure trials, significantly more than did depressed sub-
jects. As predicted, there were no significant differences between
groups on the chance task. Moreover, Beck Depression Inventory Scores
correlated negatively and significantly with total amount of expec-
tancy change on the skill task. From this data, Miller and Seligman
(1973) concluded that depressed people do not perceive response
dependent reinforcement as, in fact, response dependent. What this
implies is that the depressed person is insensitive to his own suc-
cesses.

Aaron Beck, reflecting on his clinical experience, has argued
an opposite position: that the depressed person is supersensitive to
any information about his performance. Even a minor success or fail-
ure will affect his evaluation of himself and of his likelihood of
future success. Loeb, Beck, and Diggory (1971) compared depressed and
nondepressed psychiatric patients' reactions to success or failure on

a difficult card-sorting task. Patients were assigned to either

31

depressed or nondepressed groups according to Beck Depression Inven-
tory scores and to a psychiatrist's independent rating of depth of
depression. The criterion BDI score was 22. The patient's job on
each of two similar tasks was to sort 20 cards in 20 seconds. By
varying the actual length of each trial, the experimenter controlled
each subject's performance. On the first five trials of the first
task, all subjects failed to reach criterion. On the sixth and
seventh trials only the success group subjects reached criterion.
All subjects failed on the second task. Prior to each trial, each
patient gave his estimate of his probability of success. On the
first task depressed subjects gave significantly lower estimates
than did nondepressed, yet for both depressed and nondepressed
groups, probability estimates on the second task varied with success
on the first. Success group patients gave significantly more opti-
mistic estimates than did their failure group counterparts. Yet even
under conditions of success, depressed patients gave lower estimates
of probability of success than did nondepressed. Loeb et a1. (1971)
concluded that depressed patients do respond favorably to success,
though they remain more pessimistic than nondepressed patients.

The findings of Miller and Seligman (1973) and of Loeb,
Beck, and Diggory (1971) suggest somewhat different conceptions of the
relation between depression and the perception of reinforcement. A
comparison of these two studies, though, has inherent limitations.
First, these two experiments drew their subjects from markedly differ-

ent populations. Miller and Seligman (1973) worked with college

32

students, none of whom were clinically depressed. Loeb et al.
(1971) drew their subjects from a psychiatric patient group. Their
depressed subjects were clinically depressed. Secondly, the studies
used somewhat different dependent measures. The former examined
changes in expectancy level, the latter simply looked at expectancy
levels. Miller and Seligman (1973) found that their depressed sub-
jects did not significantly alter expectancies for success given
previous success on a skill task. Moreover, this failure could not
be attributed to simple pessimism, for if the depressed students were
simply more pessimistic, this should also have manifested itself on
the chance task._ This was not the case. On the other hand, Loeb
and his colleagues (1971) found that highly depressed patients who
succeeded on the first task raised their expectancy for successful
performance on the second task. That they did so suggests that
depressed people can perceive and respond to the efficacy of their
own actions.

To view these two sets of findings as inconsistent presup-
poses that the perception of reinforcement is an all or none phenome-
non. One either does or does not perceive his response contingent
reinforcement as in fact, response contingent. This assumption is
debatable. The perception of reinforcement is probably determined
by several factors. One of these may be the individual's definition
of mastery of a task. To perceive oneself as mastering a task is to
view the reinforcements one receives as a function more of one's

developing skill than of good luck. If one's definition of mastery

33

is rigidly perfectionistic, any performance short of complete success,
fails to qualify as mastery. Thus, the more stringent one's defini-
tion, the larger the percentage reinforcements one needs in order to
accept these as the outcome of one's skill. As noted earlier, the
depression prone person often maintains rigid and perfectionistic
standards of success. He draws a false dichotomy between perfect
performance and failure. He allows himself little credit for medi-
ocre, though adequate, performance. The depressed person rarely sees
himself as effective because he maintains excessively high criteria
for mastery of various life tasks. Thus, what may differentiate
depressed from nondepressed persons is not simply that the former

do not see their reinforcements as response dependent, but rather,
that they require relatively more reinforcements to do so. If they
are reinforced on a sufficiently large percentage of task trials,
depressed persons will view their successes as response dependent and
adjust their expectancies for future success accordingly.

The hypothesis that depressed people require more positive
feedback in order to see themselves as mastering a taSk may FESOIVE
the discrepancy between the findings of Miller and Seligman (1973)
and those of Loeb et al. (1971). In the latter study, the experi-
menter controlled each patient's performance such that he better
approximated criterion on each trial. Thus, even highly depressed
subjects could perceive themselves as mastering this task. In
Miller and Seligman's (l973) experiment, the distribution of rein-

forced trials was random. This may have made it more difficult for

34

subjects, and particularly the depressed ones, to determine whether
they were mastering their task. Had the depressed subjects been
reinforced on a larger percentage of task trials, they might have
been persuaded, by this unambiguous feedback, that they were master-
ing the skill task.

The focus of the present study was to test the generality of
Miller and Seligman's (l973) findings. This was done through an
expanded replication of Miller and Seligman's (1973) procedures. The
perception of reinforcement on skill and chance tasks, on either a
50% or a 75% reinforcement schedule, was compared for depressed and
nondepressed college students. This perception, which is reflected
in changes in expectancy of success on proceeding trials, was
assessed by three dependent measures: change in expectancy level
following the first, reinforced trial, expectancy level following the
last, reinforced trial, and the sum of the absolute values of all
trial—by-trial changes in expectancy level. Seven specific hypothe-
ses were tested.

Hypothesis 1: When reinforced on 50% of the trials on the
skill task, nondepressed subjects will modify
expectancy levels significantly more than

depressed subjects, as reflected by all three
dependent measures.

 

When reinforced on only half of the skill task trials, non-
depressed subjects, but not depressed subjects, will correctly per-
ceive reinforcements as response dependent.

Hypothesis 2: When reinforced on 75% of the trials on the
skill task, depressed and nondepressed sub-
jects will not differ in modifying expectan-
cies, as reflected by the second and third
dependent measures. Nondepressed subjects,

 

35

however, will raise expectancies following
the first, successful trial significantly
more than the depressed subjects.

Depressed subjects, like nondepressed, if reinforced on a
large percentage of skill task trials, correctly perceive these suc-
cesses and response contingent and adjust their expectancies accord-
ingly. Depressed subjects, though, require more than one success in
order to perceive their rewards correctly. Thus depressed subjects
should differ from the nondepressed on the first dependent measure.

Hypothesis 3: Although reinforced at the same 75% rate,
depressed subjects will alter trial-by-trial
expectancy levels on the skill task, differ-
ently than nondepressed. On the first half
of the trials, trial-by-trial changes will be
smaller for the depressed than for the non-
depressed subjects. On the second half of
the trials, these changes will be smaller
for the nondepressed subjects.

 

Depressed subjects are slower to perceive their rewards as
response dependent. For this reason their expectancy levels should
reach asymptote later (in the task session) than those of nondepressed
subjects. Only after a few successful trials will the depressed sub-
jects begin to appreciably shift their expectancy levels.

Hypothesis 4: On the chance task, for both 50% and 75% rein-
forcement schedules, depressed and nondepressed
subjects will not differ in changing expectancy

levels, as reflected by all three dependent
measures.

 

Both depressed and nondepressed subjects correctly perceive
chance-task successes as response independent.

Hypothesis 5: Nondepressed subjects will modify expectancies
significantly more on the skill task than on the
chance task, at both 50% and 75% reinforcement
schedules.

 

36

Nondepressed subjects differentiate between response con-
tingent and response independent reinforcements, and adjust their
expectancy levels appropriately.

Hypothesis 6: When reinforced on 50% of the trials,
depressed subjects will modify expectan-
cies no differently on the skill task than
on the chance task, as reflected by all
three dependent measures.

When reinforced on only half the trials, depressed subjects (
fail to perceive skill task successes as response contingent. Thus
these subjects respond no differently to skill task successes than

to chance task successes.

 

Hypothesis 7: When reinforced on 75% of trials, depressed
subjects modify expectancies on the skill
task significantly more than on the chance
task-on final expectancy level. On the
first dependent measure, there should still
be a significant skill versus chance differ-
ence.

Depressed subjects, when sufficiently reinforced, perceive
response-reinforcement relations correctly. However, these subjects

need more than one successful trial to differentiate between tasks.

CHAPTER II

METHODS

Subjects

The 40 subjects who participated in this study were all stu-
dents attending summer session psychology classes at Michigan State
University. They were recruited from a pool of 137 students who had
completed the Beck Depression Inventory (801) and the Empathic-
Similarity Scale. This latter instrument was used as a "filler."

The subject pool was limited to students between the ages of 17 and
25. The mean 801 score for this group was 7.16, and the standard
deviation was 5.64. Students whose BDI scores were either 5 or below,
or 9 or above were subsequently invited to participate in the labora-

tory experiment phase of the study. Table 1 presents the B01 scores,

TABLE 1

Beck Depression Inventory Scores, Age and Sex
Characteristicscfl the Two Experimental Groups

 

 

Mean Mean
Group N 801 Age Sex
Nondepressed 20 2.95 20.10 12F, 8M
Depressed 20 14.45 20.45 12F, 8M

 

37

 

38

ageland sex characteristics of the nondepressed and depressed subject
groups. Within both of these groups, half of the subjects were
assigned to the 50% reinforcement condition and half to the 75% rein-
forcement condition. Half of the subjects in each of the four Depres-
sion X Reinforcement cells were assigned to the skill-chance task
order and half to the chance-skill order. For each of the depression
level groups, a separate, prespecified random order determined the
Reinforcement X Task Order cell to which each subject was assigned.
All subjects whose BDI scores qualified them for participation in the
experiment were contacted by telephone in order to set up individual
times for them to come for the experiment. These appointments were
scheduled within ten days, and in most cases within five days follow—
ing completion of the 801. All students received course credit for
both completing the Beck inventory and participating in the labora-

tory procedure.

Instruments

 

The Beck Depression Inventory is a self-report instrument
which measures depth of depression. Its construction presupposes
that this varies directly with the number and intensity of depressive
symptoms. The 801 includes 21 clinically derived categories of atti-
tudes and physical symptoms specific to depression (Beck, 1972, -

p. 189). Each category item consists of four or five graded self—
report statements. Numerical values for these range from O, for
neutral, to 3, for maximal severity. Individual scores, which range

from O to 63, are determined by the number and severity of symptoms

 

‘- _z—. .
J

39

reported. Beck (1972, p. 196) reports mean scores associated with
clinical ratings of depression for a psychiatric population. These
range from 18.7 for mildly depressed patients to 30 for severely
depressed.

The body of data on the reliability and validity of the 801 '
is impressive (Beck, 1972, pp. 184-202). The scale is internally
consistent; split half reliability is .93, and item total correla- (
tions range from .31 to .68. Concurrent validity of the inventory
has been established by its high correlations with other measures of

depression. Beck (1972, pp. 197-198) reports correlations of .65

 

and .75 between 801 scores and clinicians ratings of severity of LN“
depression. BDI scores also correlate well with scores on Lubin's
Depression Adjective Check List (r=.66) and the MMPI D-Scale (r=.75)
(Beck, 1972, p. 198). Unlike these latter two, the BDI discriminates
between depression and anxiety. Moreover, changes in 801 scores can
predict changes in clinicians' ratings (Beck, 1972, p. 198). Con-
struct validity of the inventory has been supported by confirmation

of several hypotheses. These are: Depressed patients are more likely
to report dreams with themes of deprivation, failure, and frustration,
to identify with "the loser" in projective tests dealing with success
and failure, to have a history of childhood deprivation, and to
respond to experimentally induced failure with a disproportionate

drop in self-esteem. In short, the Beck Depression Inventory is as
good a measure of depression as one can reasonably expect of a brief,

D easily administered self-report questionnaire.

40

Apparatus

Chance Task

 

The equipment used in this phase of the experiment was a
Kodak Carousel 800 slide projector. The revolving slide tray con-
tained three slides: one marked with an "X," one marked with an "O,"
and one blank. The blank slide was projected onto the wall between
presentations on the other two slides. In this way, the experimenter

controlled the sequences of "X"s and "O"s.

Skill Task

 

The Vertical Aspiration Board employed in the skill task is
the same as that used by Rotter et al. (1961). It consists of a
movable wooden platform block set within a vertical wooden frame,
just over two feet tall. The platform can be raised and lowered by
pulling on an eight foot long string which is attached to the top of
the platform and passes through a pulley. A one-half inch steel
bearing placed on the platform is actually held in place by a con-
cealed electromagnet, which is inserted into a three-fourths inch hole
in the back of the platform. Strips of spring brass are connected to
the electromagnet and attached to both sides of a wooden flange at
the back of the platform block. The interior track of the frame is
lined with brass strips. The spring brass on the flange thus remains
in contact with the strips as the platform is raised and lowered.
Concealed wires connect the brass strips to a hidden six-volt dry cell
and to a concealed silent switch. Both of these are concealed under

the table top to which the apparatus is bolted. The top of the

 

41

platform tilts forward slightly so that the bearing will roll off
the platform, unless it is held in place by the electromagnet. The
silent switch controls the electromagnet. The subject, who stood
six feet from the apparatus was required to raise the platform

almost to the top of the frame without dislodging the steel bearing.

Procedure
When each subject reported to the laboratory, the experimen-
ter first noted his 801 score and then assigned him to one of four
Reinforcement X Task Order cells, according to a prespecified random

order. The experimenter then very briefly explained the nature of

 

the two experimental tasks. Following this she read the instructions
for performing either the skill task on the chance task and for
estimating degree of certainty of success. These were:

Skill: This task has two purposes. The first is to see how
well you can succeed in raising the platform without letting the ball
fall off. The second is to see how accurate you are in estimating
your success. Your objective is to try, by pulling the string, to
raise the platform to the “90" mark without letting the ball drop off.
The apparatus tilts forward slightly so that the ball is more likely
to drop off the higher the platform is raised. Of course, if you
raise the platform very quickly, the ball cannot drop off because of
its momentum. Therefore, you should raise the platform slowly. We
will repeat this eight times. Any questions?

 

Chance: This task [also] has two purposes. The first is to
see how well you can tell me beforehand which of two types of slide
will appear on the wall. The second is to see how accurately you can
estimate your success. In this projector I have some slides marked
either with an "X" or an "O." I shall be showing you these in sets
of five. Their order of presentation will be random; that is, the
sets of five presentations have no special patterns, nor will they
necessarily have the same number of "X"s and "O"s. Your job is to
guess which slide will appear next before I present it. Each set of
5 slide presentations will count as one trial. In order to be suc-.
cessful on a trial, you must get at least four of the five slides

42

right. Any number correct, less than four will mean you have not
succeeded. We will repeat this eight times. Any questions?

Estimating Certainty of Success: Now, before each trial I
would like you to estimate how certain you are that you will raise
the platform to the "90" mark. [Get at least four of the five slides
right.] You are to estimate your degree of certainty of success on
a scale going from O to 10. For example, if you feel fairly certain
you will succeed, you may rate yourself with a high number such as
9 or 10. If you feel moderately sure that you will succeed, you may
rate yourself with a number at the middle of the scale, such as a
4, 5, or 6. If you feel pretty sure you will ngt_be successful, you
may rate yourself with a low number such as O or 1. You may use any
number from O to 10, inclusive. Bpt_it is important that you select
your estimates carefully, agg_that they correspond closely with how
certain you really are. So, before we begin,make an estimate on the
O to 10 scale as to how likely you think you are to succeed on this
first trial.

 

Each subject was given eight trials on each task. The experi-
menter controlled trial-by—trial outcomes on both. On the skill task,
she did this by turning the switch which controlled the electromagnet.
On failure trials, she would turn off the magnet before the platform
block reached the "90" mark on the vertical frame; on success trials
the bearing was held in place until the criterion was reached. Suc-
cess and failures on the chance task were manipulated by controlling
the presentation of "X"s and "O"s so that these might either coincide
or disagree with the subject's predictions. Right and wrong guesses
within each trial of five presentations were randomized. The experi-
menter allowed the subject either two or three correct guesses on
failure trials. On success trials the subject was allowed four cor-

rect guesses. The order of reinforced and nonreinforced trials was:

50% + - _ — .+ + - - +

75% + + — + + — + +

 

43

No reinforcer other than the knowledge of having succeeded was
employed.

Several subjects either recognized how the experimenter con-
trolled the Vertical Aspiration Board, or jerked the pulley string
too quickly for her to control the outcome of each trial. Their data
was discarded, and replaced by conducting the procedure for additional
subjects.

After completing both tasks, subjects were de-briefed and

thanked for their participation.

Dependent Measures

 

Three dependent measures assessed the effects of task type,
level of depression, and amount of reinforcement, on verbalized
expectancy of success. The first measure was change in expectancy
level following the first, reinforced trial, that is, the difference
between expectancy prior to Trial 2 and expectancy prior to Trial 1.
The other two measures assessed the cumulative effects of successes
and failures. These were expectancy following the last trial and
the sum of the absolute values of all increases following success

and decreases following failure.

 

CHAPTER III

RESULTS

In order to assess the effects of both level of depression
and rate of reinforcement on changes in expectancy level over task
trials, a Depression X Task X Reinforcement analysis of variance,
with repeated measures on the Task factor was computed for each of
the three dependent measures. These analyses assessed both the over-
all Depression X Task interaction and specific Depression X Task X
Reinforcement interactions. The specifically hypothesized differ-
ences between depressed and nondepressed subjects on each task were
tested by simple t:tests.

A between groups comparison of changes in level is unambigu-
ous only if these groups are equivalent on initial expectancy level.
Thus a preliminary Depression X Task X Reinforcement ANOVA, with
repeated measures on the Task factor,was computed for verbalized
expectancies given just prior to the first trial on each task. This
yielded no significant main effects and no significant interactions
on Task X Depression, Depression X Reinforcement, and Depression X
Task X Reinforcement (all ps > .05). A Task X Reinforcement inter-
action was significant at the .05 level. A post hoc examination

[Tukey test] of this interaction indicated that the 75% reinforcement

44

 

JTT

45

group subjects gave higher estimates of succeeding on the first skill
task trial than on the first chance task trial. The 50% Reinforce-
ment group subjects responded no differently to the skill task than
to the chance task. How or why rate of reinforcement should affect
expectancy of success prior to any experience with the task is not
obvious. One possible explanation of this Task X Reinforcement
interaction is that it is actually due to a task-order effect. A
complex Latin Square design analysis of variance, however, revealed

no significant order effects. The only remaining factor to account

 

for this unexpected interaction is random subject variability: it
just happened that the 75% reinforcement group subjects initially
responded differently to the two tasks. While this initial differ-
ence must be considered in any further interpretation of the data, it
does not seem a significant obstacle to further data analyses.

Table 2 presents the mean scores on the three dependent meas-
ures, on skill and chance tasks for each of the four groups. Tables
3, 4, and 5 present the results of overall £_tests and t_tests com-
puted for each of the dependent measures. Overall, these data support
neither the Task X Depression interaction predicted from the learned
helplessness theory nor the variation in this interaction according
to reinforcement level as hypothesized by this author. There is,
however, some evidence for a Depression X Task interaction in a
direction opposite to that predicted. Depressed subjects differed
from nondepressed in adjusting expectancies on the chance task,

rather than on the skill task.

TABLE 2

Means for Each of the Four Groups on
Skill and Chance Tasks

46

 

 

 

 

 

 

 

 

Nondepressed Depressed r
Measure 50% 75% 50% 75% ['"F
S C S C S C S
Change from
Trial 1 to Trial 2 4.2 1.0 2.8 1.3 4.3 1.5 2.8 2.1
Expectancy after '
Final Trial 9.3 8.5 9.7 6.2 8.5 6.1 9.7 8.0
Total Amount of
Expectancy Change 9.6 4.6 6.0 5.2 11.4 6.8 7.2 8.3

 

47

TABLE 3

Analysis of Variance of Change in Expectancy
from Trial 1 to Trial 2

 

 

 

 

 

 

Source df ms .5
Between 39
Depression 1 2.45 <1
Reinforcement l 5.00 1.47
D x Rf 1 .50 <1
Error 36 3.39
Within 40
Tasks l 84.05 42.67***
T X D l 1.80 <1
T X Rf l 18.05 9.11**
T X D X Rf l .25 <l
Error 36 1.98 <1
tftests

A. Nondepressed vs. Depressed 3. (df = 1/36)

50% Skill 1

75% Skill 1

50% Chance 1

75% Chance 1.27
B. Skill vs. Chance

50% Nondepressed 5.08***

75% Nondepressed 2.38*

50% Depressed 4.45***

75% Depressed 1.11

fp.< .05

** < .01

Hm}.m1

(7

TABLE 4

Analysis of Variance of Expectancy Following the
Final Trial

48

 

 

 

 

 

Source df MS .5
Between 39
Depression 1 3.20 <1
Reinforcement 1 22.05 5.02*
D X RF 1 5.00 1.11
Error 36 4.39
N-lthln 40
735k 1 162.45 92.30****
T X D 1 12.80 7.27**
T X Rf 1 1.25 <1
T X D X Rf 1 .20 <1
Error 36 1.76
tftests
A. Nondepressed vs. Depressed t_ (df = 35)
50% Skill . 1,35
75% Skill 0
50% Chance 1,0]
75% Chance 3,o3***
B. Skill vs. Chance
50% Nondepressed 6,40****
75% Nondepressed 5,90****
50% Depressed 4,03****
75% Depressed 2.86***
fp < .05
*fp < .025
*** < .01

****_p_ < .001

 

TABLE 5

Analysis of Variance of Total Amount of Expectancy Change

49

 

 

 

 

 

Source df MS .5
Between 39
Depression 1 86.13 3.92
Reinforcement l 40.63 1.8
D X Rf 1 12.50 <1
Error 36 21.95
Within 40
Task 1 108.13 10.29**
T X D 1 6.60 <1
T X Rf l 68.50 6.52*
T X D X Rf l 44.43 4.23*
Error 36 10.50
t7tests (df = 36)

A. Nondepressed vs. Depressed t

50% Skill 1.24

75% Skill <1

50% Chance 1.55

75% Chance 2.13*
B. Skill vs. Chance

50% Nondepressed 3.45**

75% Nondepressed <1

50% Depressed 2.48*

75% Depressed l

 

*p_< .05
**p_< .Ol

50

The results of this experiment do not support the first
hypothesis that when reinforced on 50% of the skill task trials,
nondepressed subjects will change expectancies for success more than
depressed subjects. Simple inspection of the mean changes in expec-
tancy following the first trial shows that these groups did not
differ. The t_test comparison of mean final expectancies, skill task,
50% reinforcement, presented in Table 4, indicated no significant
differences between the depressed and nondepressed groups. Similarly,
these groups did not differ on the third dependent measure: total
amount of expectancy change (see Table 5).

The second specific hypothesis tested was that nondepressed
and depressed subjects, when reinforced on 75% of the skill task
trials, would not differ with regard to final expectancy and total
expectancy change. With regard to change following the first, suc-
cessful trial, nondepressed subjects would still show larger changes
than depressed subjects. This hypothesis was supported for the most
part. As can be seen from Tables 4 and 5, t_tests revealed no sig-
nificant differences between groups (both t_< 1, ps > .05). On the
first dependent measure, on which a significant difference was
expected, there was no difference between depressed and nondepressed
(Table 4, t_< l, p_> .05).

The third hypothesis stated: when reinforced on 75% of the
skill task trials, nondepressed and depressed groups would manifest
different patterns in modifying their expectancy levels. The non-

depressed subjects were expected to modify expectancies more on the

 

51

first few trials and to reach asymptote sooner than the depressed
subjects. The latter were expected to shift expectancies less on the
first several trials and to then compensate for this on the later
trials. An analysis of variance, presented in Table 6, of the eight
changes in expectancy from trial to trial showed no interaction

between depression and task trials.

TABLE 6

Analysis of Variance of Trial by Trial Changes
in Expectancy on the Skill Task - 75% Rf

 

 

 

 

Source df MS F
Between 19

Depression 1 .90 1
Error 18 1.44

Within 140

Trials 7 14.05 15*
Trials x Depression 7 .20 <1
Error 126 .9

fp < .001

Hypothesis 4: 0n the chance task, at both 50% and 75% rein-
forcement, nondepressed and depressed subjects would not differ in
adjusting their expectancies; was not confirmed. As can be seen from
Table 3, nondepressed and depressed subjects did not differ on the
first dependent measure. On final expectancy on the chance task,

there were significant differences between these groups. A EftESt

52

comparison of the 75% reinforcement groups, (Table 4) showed that
depressed subjects verbalized significantly higher final expectancies
than did their nondepressed counterparts. The overall f_test for

this measure yielded a significant Depression X Task interaction.
Analysis of the simple main effects of depression showed no differ-
ence between groups on the skill task and a significantly higher mean
final expectancy on the chance task for the depressed groups (£_= 8.5,
.2 < .01, df 1/36). On the third measure, total amount of expectancy
change, the nondepressed group did not differ from the depressed at
50% reinforcement (see Table 5). The depressed, 75% reinforcement
group subjects, however, manifested significantly larger total changes
in expectancies on the chance task than did their nondepressed
counterparts.

The fifth hypothesis tested was that nondepressed subjects,
in both reinforcement groups, would change expectancies on the skill
task significantly more than on the chance task. This was confirmed.
When reinforced on 50% of task trials, these subjects raised skill
task expectancies significantly more on all three dependent measures
(all ps < .01) (see Tables 3, 4, and 5). For the 75% reinforcement
group subjects the skill-chance difference was significant on the
first two dependent measures (both 95 < .05). 'The third measure,
total expectancy chance, did not reflect this difference. This dis-
crepancy is consistent with findings of Rotter et a1. (1961) that
under conditions of high reinforcement skill-chance differences on
this measure wash out because verbalized expectancies tend to reach

asymptote on final, consistently reinforced trials.

 

subj
the
the

trial

tri
lat

1'63

53

Hypothesis 6: When reinforced on 50% of the trials, depressed
subjects will modify expectancies no differently on the skill than on
the chance task; was not confirmed. On all three dependent measures,
these subjects modified skill task expectancies significantly more
than chance task expectancies (all ps < .01).

The final hypothesis: depressed subjects, reinforced on 75%
of task trials would show a significant skill-chance difference on ..e.
final expectancy, but not on change following the first reinforced {
trial, was confirmed. As can be seen from the t_values presented on

Tables 3, 4, and 5, only on the second measure did these subjects

 

respond differently to the two tasks. I

CHAPTER IV

DISCUSSION

The guiding premise of this study was that depressed people
differ from nondepressed in their perception of response contingent ‘
reinforcement. Learned helplessness theory, as presented and sup-
ported in Miller and Seligman's (1973) paper, argues that depressed I

individuals are impervious to the relation between their actions and

 

the rewards these yield. Clinical and experimental evidence pre- 5—
sented by Beck and his colleagues (Beck, 1972, Loeb et a1., 1971)
suggests an opposite view: depressed patients are actually hyper-
sensitive to any feedback about their performances on tasks. In an
attempt to reconcile these apparently disparate views, this author

has suggested a somewhat different approach to the perception of rein-
forcement. The perception of reinforcement as either response depend-
ent or response independent is not an all or none phenomenon, but
varies according to several factors. Two of these factors are the
individual's definition of mastery of a task and the rate at which

his efforts are rewarded. The more stringent the person's definition
of mastery, the more successes he needs in order to see these as
dependent more on his skill than on his luck. As Beck has noted
(1972, p. 7), depression prone persons maintain rigid and perfection-

istic standards for themselves. The main hypothesis of the present

54

55

study, thus, is that depressed persons require a higher rate of rein-
forcement in order to correctly perceive these rewards as response
dependent. Clearly, support for this hypothesis would limit the
generality of Miller and Seligman's (1973) conclusion.

Any test of the generality of an experimental finding
requires a replication of it as well as a test of it under different
conditions. The present study failed to replicate Miller and Selig-
man's (l973) findings. The comparison of the 50% reinforcement group
subjects yielded no difference between depressed and nondepressed
subjects in modifying their expectancies for success on the skill
task. Moreover, contrary to prediction, depressed subjects, as well
as nondepressed, manifested significantly larger changes in expectancy
on the skill task than on the chance task. AS predicted, there were
no significant differences between the depressed and nondepressed
groups who were successful on 75% of the skill task trials. Obvi-
ously, this finding lacks the import it would have had, were there
the anticipated differences between groups in the 50% reinforcement
condition. The results of this experiment indicate that the
depressed subjects perceived and responded to their successes on
the skill task as appropriately as did the nondepressed subjects.

In short, these findings support neither the learned helplessness
theory nor the reformulation of it this author proposed.

Had the results of this experiment simply shown that depressed
subjects did not differ from nondepressed in their perception of

skill task reinforcements, this discussion would conclude with a

5- ‘5'0 Ami“

:\ A «V “‘5'” f.’

 

56

recommendation for additional experimentation. This is because the
results of this one experiment can no more invalidate learned help-
lessness theory than Miller and Seligman's (1973) results can vali-
date it. Yet this study yielded another significant and wholly
unexpected result: the depressed subjects differed from the non-
depressed in modifying expectancies on the chance task.

The depressed subjects who were reinforced on 75% of the
chance task trials produced higher final expectancies and larger
amounts of total expectancy change than did their nondepressed
counterparts. This finding suggests that these groups differed in
their perception of chance task successes. The depressed subjects
were far more encouraged by their successes on this task than were
_the nondepressed subjects. With each success, the depressed subjects
grew even more certain of succeeding again. Such an increase in
certainty is a totally illogical response to a chance task. Anyone
who comprehends the nature of chance events knows their outcomes are
independent of each other. Thus, the most reasonable strategy for
estimating one's probability of succeeding on chance task trials is
to pick one estimate for all task trials. To modify one's expectan-
cies according to prior successes and failures is to misinterpret
these outcomes. In this sense, both the depressed and nondepressed
subjects misinterpreted their successes on the chance task; however,
the former were even more seriously mistaken in their response to
this success. The depressed subjects responded to the chance task

in a manner actually appropriate to the skill task. This suggests

- T'TR’!

 

57

that these subjects attributed their consistent success on the chance
task to their own skillful intuition.

This conclusion is opposite to that drawn by Miller and Selig-
man (1973). While the latter study concluded that depressed subjects
respond to both skill and chance task successes as if these were
response independent, the present study suggests that these subjects
respond to both types of success as if they were response dependent.
These opposite findings present a paradox. There are two ways out of
a paradox. One way is to assume that one of the conclusions is false
and that further research will demonstrate this. The other way is to
search for a wider conceptual base from which the paradox can be
resolved. Further consideration of all of the available data,
together with our current understanding of the cognitive aspects of
depression points to a broader conceptualization of the relation
between depression and the perception of reinforcement.

This reformulation assumes that the dichotomy between response
dependent and response independent reinforcement applies almost
uniquely to the experimental laboratory. In "real life" one's suc-
cesses and failures are rarely a function solely of skill or of chance.
Rather, most events are determined both by one's skill and by factors
beyond one's control. A reasonable, healthy reaction to one's cir-
cumstances presupposes a realistic assessment of the roles played by
one's own efforts and by other, uncontrollable factors. The well-
adjusted, nondepressed person can maintain such an assessment and

can take his gains and losses pretty well in stride. The depressed

 

58

person, however, maintains a distorted view of what he can and cannot
control. He may feel helpless, but also responsible for events he
could not possibly determine. Beck (1972, p. 232) offers an extreme
example of this in the case of the depressed housewife who, when she
took her children on a picnic and it stormed, blamed herself for not
picking a better day. Seligman (1974, p. 95) emphasizes that the
critical feature of depression is the perceived loss of control of
one's punishments and rewards. Yet Seligman has apparently divided
rewards into response dependent and response independent and referred
this perceived loss of control only to the former. This may be too
narrow a premise.

A more comprehensive premise is that the depressed person has
a faulty notion of the roles played both by his own skillful effort
and by factors external to himself in shaping his predicament. Thus
he may either underestimate or overestimate the effects of his own
behavior on his situation. He may either fail to recognize the
rewards of his efforts or blame or credit himself or events which
were not of his doing. In short, the depressed individual may mis-
perceive either primarily response dependent or primarily response
independent rewards, or both. This reformulation of the relation
between depression and the perception of reinforcement may reconcile
the results of the present study with those of Miller and Seligman
(1973).

The hypothesis that depressed individuals may misinterpret

response dependent and/or response independent reinforcements cannot

 

59

by itself, explain the discrepancy between Miller and Seligman's
(1973) results and those of the present study. Additional factors
are necessary to explain why these particular two groups of depressed
subjects perceived their skill and chance task successes in opposite»
ways. There are two possible reasons for this difference. The first
possibility is that the skill and chance tasks actually differed
significantly between the two studies. The other possibility is that
the depressed subjects differend between the two studies. Both of
these explanations are, however, mere speculations.

There were minor differences between the ways these two stud-
ies conducted the experimental tasks. In the present study, the
experimenter timed skill task failures (i.e., turned off the "silent
switch" so that the steel bearing rolled off the platform block) so
that the subjects almost succeeded on those trials. In this way
the skill task may have appeared more controllable to the depressed
subjects than the task did to Miller and Seligman's (l973) depressed
subjects. This may explain why the former, but not the latter, group
modified their skill task expectancies appropriately. The.chance task
in the current study may also have appeared somewhat different from
that in Miller and Seligman's (l973) experiment. In the present
experiment the presentation of each "X" and "0" was preceeded by a
blank slide. Thus, the sequence of "S"s and "O"s may not have
appeared truly random to the subjects in that study. While this
might account for the behavior of the depressed subjects in the 75%

reinforcement group, it cannot account for the chance-task

 

60

appropriate behavior of all of the 50% reinforcement group and of
the nondepressed 75% reinforcement group subjects.

A second possibility is that the depressed subjects in these
two experiments were, in fact, different. The depressed subjects in
Miller and Seligman's (1973) study were given the Beck Depression

Inventory immediately prior to the experimental procedure. In the

current study, however, several days elapsed between these two parts
of the study. It might be that, because of this passage of time, the

depressed subjects were no longer so depressed at the time of the

Fm‘fim‘. imam: ante-Isl Inns-fl
I

laboratory procedure. This possibility can account for this study's

1

failure to replicate Miller and Seligman's (1973) results, but not
for the difference between depressed and nondepressed subjects on the
chance task.

The upshot of this study is that the relation between depres-
sion and the perception of reinforcement remains unclear. The hypothe-
sis that depressed individuals misconstrue both response dependent
and response independent reinforcements is worth investigating. This
hypothesis presupposes that the depressed person maintains a faulty
notion of which events he can and cannot control by his own efforts.
This presupposition might be tested by comparing depressed and non-
depressed subjects' responses to a task in which the subjects'
rewards are contingent both on his skill and on factors outside of
his control. For example, the Vertical Aspiration Board skill task
could be varied so that the subject receives monetary rewards for

his "successes" on a random basis, or a puzzle or anagram task could

Name: Class: _F

 

Age: Sex: Telephone:

 

EMPATHIC - SIMILARITY SCALE

Instructions: Please read each statement and circle T or F as it applies
to you.

1. I like to have people around me practically all of the time.

2. I always prefer to work with others.

3. I would rather listen to a story than tell one.

4. I am cautious about undertaking anything which may lead to
humiliating experiences.

5. I am almost never embarrassed.

6. I am fairly easily moved to laughter or tears.

7. I am moderate in my tastes and sentiments.

8

9

FBI-31814.1
*4’11'11’121

. I think much and speak little.
. I have strong likes and dislikes.

10. It takes a great deal to make me emotional.

11. I become emotional fairly easily.

12. I have sometimes corrected others, not because they were

wrong, but only because they irritated me.

13. I am mainly interested in ideas that are very practical.

. I get an intense pleasure from just looking at a beautiful
building.

Fatal-396881316
"d’flfi'fi'fi'fl'flfi

Iii-3
iaiq
p
:—

T F 15. I generally seek whatever makes me happy here and now.

T F 16. I always keep control of myself in an emergency situation.

T F 17. I find it rather hard to keep to a rigid routine.

T F 18. It is necessary to retain the belief that God exists as a
personal being.

T F 19. In matters of conduct I conform very closely to custom.

T F 20. The thought of God gives me a complete sense of security.

T F 21. I control my sexual impulses by instituting prohibitions and
restrictions.

T F 22. Some of my friends think my ideas are a bit wild and impractical.

T F 23. I am temperamentally more a sceptic than a believer.

T F 24. The European attitude toward mistresses is more sensible than

ours.

61

be devised which can be rigged so that the subject can solve it only
on certain randomly selected trials. The subject, in either case,
would be told that his own efforts would not necessarily insure a
reward. If depressed and nondepressed subjects' changes expectancy
of success on this task, as well as the usual skill and chance
tasks, under low, medium, and high reinforcement rate conditions,
were compared, a clearer picture of the relationship between depress

sion and the perception of reinforcement might emerge.

APPENDIX

SCALES OF MEASUREMENT

62

Name: _¥ hp Telephone #:

 

 

Age:

 

Sex:

 

BECK'S D.I.

Instructions: Please read each get of statements completely, then
circle the number of the one which most represents how you feel
right p93, For example, read all_the statements in category ”A",
reflect for a minute, then choose one of them and circle it. Then
continue to the next set until you have chosen one statement for
every letter through "U".

A. I do not feel sad

I feel blue or sad

I am blue or sad all the time and I can't snap out of it

I am so sad or unhappy that it is quite painful

I am so sad or unhappy that I can't stand it
B. I am not particularly pessimistic or discouraged about the future

I feel discouraged about the future

I feel I have nothing to look forward to

I feel that I won't ever get over my troubles

I feel that the future is hopeless and that things cannot improve
C. I do not feel like a failure V

I feel I have failed more than the average person

I feel I have accomplished very little that is worthwhile or

that means anything

As I look back on my life all I can see is a lot of failures

I feel I am a complete failure as a person (parent, husband, wife)
D. I am not particularly dissatisfied

I feel bored most of the time

I don't enjoy things the way I used to

I don't get satisfaction out of anything any more

I am dissatisfied with everything
E. I don't feel particularly guilty

I feel bad or unworthy a good part of the time

I feel quite guilty

I feel bad or unworthy practically all the time now

I feel as though I am very bad or worthless
F. don't feel I am being punished

have a feeling that something bad may happen to me
feel I am.being punished or will be punished

feel I deserve to be punished

want to be punished

HHHHH

 

M.

N.

hikihiki Fibihl Fihihlhl 14141414 141414141414 14141414 1414141414

hlhiki

I

I

don't feel disappointed in myself
am disappointed in myself

don't like myself

am disgusted with myself

hate myself

don't feel I am any worse than anybody else

am critical of myself for my weaknesses or mistakes
blame myself for my faults

blame myself for everything had that happens

don't have any thoughts of harming myself

have thoughts of harming myself but I would not carry them out
feel I would be better off dead

feel my family would be better off if I were dead

have definite plans about committing suicide

would kill myself if I could

don't cry any more than usual

cry more now than I used to

cry all the time now. I can't stop it

used to be able to cry but now I can't cry at all even though
I want to

am.no more irritated now than I ever am

get annoyed or irritated more easily than I used to

feel irritated all the time

don't get irritated at all at the things that used to irritate me.

have not lost interest in other people

am less interested in other people now than I used to be

have lost most of my interest in other people and have little
feeling for them

have lost all my interest in other people and don't care about
them at all

make decisions about as well as ever

try to put off making decisions

have great difficulty in making decisions
can't make any decisions at all any more

don't feel I look any worse than I used to

am worried that I am looking old or unattractive

feel that there are permanent changes in my appearance and they
make me look unattractive

feel that I am ugly or repulsive looking

can work about as well as before

It takes extra effort to get started at doing something

I
I
I

don't work as well as I used to
have to push myself very hard to do anything
can't do any work at all

 

*3

H HHH

hibikikl

can sleep as well as usual
wake up more tired in the morning that I used to
wake up l-2 hours earlier than usual and find it hard to
get back to sleep
wake up early every day and can't get more than 5 hours sleep

don't get any more tired than usual
get tired more easily than I used to
get tired from doing anything

get too tired to do anything

My appetite is no worse than usual
My appetite is not as good as it used to be
My appetite is much worse now

I

Flhlki hihihihi

H

14141414

have no appetite at all any more

haven't lost much weight, if any, lately
have lost more than 5 pounds

have lost more than 10 pounds

have lost more than 15 pounds

am no more concerned about my health than usual

am concerned about aches and pains or upset stomach or constipation
am so concerned with how I feel or what I feel that it's hard

to think of much else

am completely absorbed in what I feel

have not noticed any recent change in my interest in sex
am less interested in sex than I used to be

am.much less interested in sex now

have lost interest in sex completely

Name: Class:

 

Age: Sex: Telephone:

 

EMPATHIC - SIMILARITY SCALE

Instructions: Please read each statement and circle T or F as it applies
to you.

1. I like to have people around me practically all of the time.

2. I always prefer to work with others.

3. I would rather listen to a story than tell one.

A. I am cautious about undertaking anything which may lead to
humiliating experiences.

5. I am almost never embarrassed.

6. I am fairly easily moved to laughter or tears.

7. I am moderate in my tastes and sentiments.

8

9

Fatal-314
“1’11'11'11

. I think much and speak little.
. I have strong likes and dislikes.
10. It takes a great deal to make me emotional.
11. I become emotional fairly easily.
12. I have sometimes corrected others, not because they were
wrong, but only because they irritated me.

ear-366161316
“11*4'111'11'11'21'11'11

T F 13. I am mainly interested in ideas that are very practical.

T F 14. I get an intense pleasure from just looking at a beautiful
building.

T F 15. I generally seek whatever makes me happy here and now.

T F 16. I always keep control of myself in an emergency situation.

T F 17. I find it rather hard to keep to a rigid routine.

T F 18. It is necessary to retain the belief that God exists as a
personal being.

T F 19. In matters of conduct I conform very closely to custom.

T F 20. The thought of God gives me a complete sense of security.

T F 21. I control my sexual impulses by instituting prohibitions and
restrictions.

T F 22. Some of my friends think my ideas are a bit wild and impractical.

T F 23. I am temperamentally more a sceptic than a believer.

T F 24. The European attitude toward mistresses is more sensible than

OUI‘S.

REFERENCES

67

REFERENCES

Beck, A. T. Cognition, affect, and psychopathology. Archives of
General Psychiatry, 1971, 24, 495-500.

 

 

Beck, A. T. Depression: Causes and Treatment. Philadelphia:
University of Pennsylvania Press, 1972.

 

Beck, A. T. The development of depression. In R. J. Friedman &
M. M. Katz (eds.), The Psychology of Depression: Contempor-
a:y_Theory and Research. New York: John Wiley & Sons,
1974, 3-21.

 

 

Becker, J. Depression: Theory_and Research. New York: John Wiley
& Sons, 1974.

Burgess, E. The modification of depressive behavior. In R. Rubin
& C. Franks (eds.), Advances in Behavior Therapy. New York:
Academic Press, 1968.

Ferster, C. B. Behavioral approaches to depression. In R. J. Fried-
man & M. M. Katz (eds.), The Psychology of Depression: Con-
temporary Theory and Research. New York: John Wiley 8 Sons,
1974: 29‘55.

 

Hiroto, D. S. Locus of control and learned helplessness. Journal
of Experimental Psychology, 1974, 192, 187-193.

Lewishon, P. M. A Behavioral approach to depression. In R. J. Fried-
man & M. M. Katz (eds.), The Psychology of Depression:
Contemporary Theory and Research. New York: John Wiley &
Sons, 1974, 157-179.

Liberman, R. P., & Raskin, D. E. Depression: A Behavioral formula-
tion. Archives of General P§ychiatry, 1971, 24, 515-523.

Libet, J. & Lewishohn, P. M. The concept of social skill with special
reference to the behavior of depressed persons. Journal of
Consulting_and Clinical Psychology, 1973, 49, 304-312.

Loeb, A., Beck, A. T., 8 Diggory, J. Differential effects of success

and failure on depressed and nondepressed patients. Journal
of Nervous and Mental Disease, 1971, l§2, 106-114.

68

69

Maier, S. F. Failure to escape traumatic shock: Incompatible motor
responses or learned helplessness? Learning and Motivation,
1970, 1, 157-170.

 

Melges, F. T., & Bowlby, J. Types of hopelessness in psychopathologi-
cal process. Archives of General Psychiatry, 1969, 29,
690-699.

 

Miller, W. R., & Seligman, M. E. Depression and the perception of
reinforcement. Journal of Abnormal Psychology, 1973, 22,
62-73.

Overmier, J. B., & Seligman, M. E. Effects of inescapable shock on
subsequent escape and avoidance learning. Journal of Com-
parative and Physiological Esychology, 1967, 42, 28-33.

Phares, E. J. Expectancy change in chance and skill situations.
Journal of Abnormal and Social Psychology, 1957, pg, 339-
342.

Rotter, J. B., Liverant, S., & Crowne, D. P. The growth and extinc-
tion of expectancies in chance controlled and skilled tasks.
Journal of Psychology) 1961, §2, 161-177.

 

Seligman, M. E. Learned helplessness. Annual Review of Medicine,
1972’ 2;, 407-412.

Seligman, M. E. Depression and learned helplessness. In R. J.
Friedman & M. M. Katz (eds.), The Psychology of Depression:
Contemporary Theory and Research. New York: John Wiley &
Sons, 1974, 83-114.

Seligman, M. E. & Maier, S. F. Failure to escape traumatic shock.
Journal of Experimental Psychology, 1967, 24, 1-9.

Seligman, M. E., Maier, S. F., & Geer, J. H. Alleviation of learned
helplessness in the dog. Journal of Abnormal Psychology,
1968, Z}; 256-262.

Thornton, J. W. & Jacobs, P. D. Learned helplessness in human sub-
jects. Journal of Experimental Psychology, 1971, 81,
367-372.

 

I‘f
i) (I, ‘l’lsl

HICHIGRN STATE UNIV. LIBRRRIES

llllll ill ill 11111 1 l
9

312 3100657919