You are on page 1of 7

Journal of Personality and Social Psychology Copyright 1984 by the

1984, Vol. 46, No. 3, 681-687 American Psychological Association, Inc.

Attributional Style and the Generality of Learned Helplessness


Lauren B. Alloy Christopher Peterson
Northwestern University Virginia Polytechnic Institute
and State University
Lyn Y. Abramson Martin E. P. Seligman
University of Wisconsin—Madison University of Pennsylvania

According to the logic of the attribution reformulation of learned helplessness, the


interaction of two factors influences whether helplessness experienced in one situation
will transfer to a new situation. The model predicts that people who exhibit a style
of attributing negative outcomes to global factors will show helplessness deficits in
new situations that are either similar or dissimilar to the original situation in which
they were helpless. In contrast, people who exhibit a style of attributing negative
outcomes to only specific factors will show helplessness deficits in situations that
are similar, but not dissimilar, to the original situation in which they were helpless.
To test these predictions, we conducted two studies in which undergraduates with
either a global or specific attributional style for negative outcomes were given one
of three pretreatments in the typical helplessness triadic design: controllable bursts
of noise, uncontrollable bursts of noise, or no noise. In Experiment 1, students
were tested for helplessness deficits in a test situation similar to the pretreatment
setting, whereas in Experiment 2, they were tested in a test situation dissimilar to
the pretreatment setting. The findings were consistent with predictions of the re-
formulated helplessness theory.

According to learned helplessness theory, situation will transfer to a new situation. These
people exposed to uncontrollable events learn factors are (a) the causal attribution the person
that their responses and outcomes are inde- makes about the uncontrollable events expe-
pendent of each other. This learning can lead rienced in the original situation and (b) the
to an expectation that responses will be futile similarity of the new situation to the original
and can generalize to new situations to inter- situation (Pasahow, West, & Boroto, 1982). If
fere with future learning (Seligman, 1975). the uncontrollable events are attributed to
Recent laboratory investigations have shown global causes, factors present in many situa-
that induced helplessness is sometimes general tions, then the symptoms of helplessness will
(e.g., Hiroto & Seligman, 1975) and sometimes generalize widely across situations. In contrast,
circumscribed (e.g., Cole & Coyne, 1977). if the uncontrollable events are attributed to
Why does learned helplessness sometimes specific causes, factors present in only a few
generalize across situations but at other times situations, then the symptoms of helplessness
remain specific to one situation? The logic of will not generalize as widely. This analysis sug-
the reformulated model of learned helplessness gests that helplessness will be relatively likely
(Abramson, Seligman, & Teasdale, 1978) sug- to generalize from the orginal situation to a
gests that the interaction of two factors influ- new similar situation, regardless of what at-
ences whether helplessness experienced in one tribution the person makes for the uncon-
trollable events in the original situation. Al-
ternatively, helplessness should be more likely
This research was supported by a grant from the National to generalize from the original situation to a
Institute of Mental Health (MH-19604) to Martin E, P. new dissimilar situation when the person
Seligman. We wish to thank the various participants in makes a global, rather than a specific, attri-
the Helplessness Seminar at the University of Pennsylvania bution for the uncontrollable events in the
for assistance in conducting the research. original situation.
Requests for reprints should be sent to Lauren B. Alloy,
Department of Psychology, Northwestern University, 1859 One factor that influences whether causal
Sheridan Road, Evanston, Illinois 60201. attributions are specific or global is the par-
681
682 ALLOY, PETERSON, ABRAMSON, AND SELIGMAN

ticular characteristics of the situation in which events. Also in both experiments, subjects were
helplessness is induced. Several laboratory then given one of the three pretreatments of
studies have shown that situational manipu- the helplessness triadic design (cf. Maier &
lations of attributions yield results consistent Seligman, 1976; Seligman, 1975): controllable
with the reformulation (see reviews of these bursts of noise, uncontrollable bursts of noise,
studies in Abramson et al., 1978, and Miller or no noise. Subjects in the uncontrollable-
& Norman, 1979). However, such demonstra- noise groups were yoked to subjects in the
tions do not explain individual differences in controllable-noise groups, receiving identical
the selection of a causal attribution in a par- bursts of noise, the only difference being that
ticular situation. the controllable-noise subjects could control
A second influence on what attributions an the offset of the noise, by pushing a button.
individual makes may be his or her general In addition, the pretreatment task was designed
attributional style (Seligman, Abramson, to be sufficiently ambiguous with regard to
Semmel, & von Baeyer, 1979). In situations the cause of a subject's failure so that attri-
in which informational cues about the causes butional styles could operate.
of events are sufficiently ambiguous, individ- In Experiment 1, subjects were tested, after
uals' attributional styles will influence their pretreatment, with a hand shuttlebox in which
causal attributions. Thus, when situational moving a lever to one side or the other of the
ambiguity about the cause of an uncontrollable box terminated bursts of noise. This test sit-
event exists, the individual who habitually uation is relatively similar to the pretreatment
makes global attributions for negative events situation, because both involve stopping an
will show more general helplessness following unpleasant noise and involve the same ex-
experience with uncontrollable events than will perimenter in the same room as part of the
the individual with a more specific attribu- same experiment. In Experiment 2, subjects
tional style. Seligman et al. (1979) described were tested with a series of anagrams. In ad-
an Attributional Style Questionnaire (ASQ) dition, the anagram task was conducted by a
for measuring individuals' dispositions to different experimenter in a different room as
make particular attributions (see also Peterson, part of a different experiment (cf. Roth & Ku-
et al., 1982). As yet, however, no study has bal, 1975). Thus in Experiment 2, the test
investigated how attributional style influences situation was designed to be highly dissimilar
the generality of laboratory-induced helpless- to the pretreatment in order to provide a rel-
ness. atively strong test of the reformulated help-
We now report two such investigations. In lessness theory. Table 1 presents the overall
both, college students responded to the ASQ design of the two experiments.
and were divided into two groups based on According to the helplessness reformulation,
mean globality-specificity scores for negative subjects with either global or specific attri-

Table 1
Overall Design of Experiments
Experiment 1: Similar test Experiment 2: Dissimilar test

Attributional Attributional
style Pretreatment Test style Pretreatment Test

global escapable noise shuttlebox global escapable noise anagrams


specific escapable noise shuttlebox specific escapable noise anagrams
global inescapable noise shuttlebox global inescapable noise anagrams
specific inescapable noise shuttlebox specific inescapable noise anagrams
global no noise (control) shuttlebox global no noise (control) anagrams
specific no noise (control) shuttlebox specific no noise (control) anagrams

Note. Experiments 1 and 2 are between-subjects designs. Attributional style was determined by the Attributional Style
Questionnaire (Seligman et al., 1979).
ATTRIBUTIONAL STYLE AND LEARNED HELPLESSNESS 683

butional styles for negative outcomes should ficient alpha, ranging between .44 and .69. For the globality
show escape deficits following experience with dimension for negative events, the dimension used in the
present investigation, alpha was .69. Test-retest correlations
uncontrollable noise in Experiment 1, because over 5 weeks ranged from .57 to .70 and for globality for
the test situation is similar to the pretreatment. negative events, it was .57.
In contrast, in Experiment 2, only subjects Subjects in Experiment 1 were divided at the median
with a global attributional style for negative of globality ratings for negative events (4.00). Subjects above
outcomes should show anagram deficits fol- the median were considered to have a global attributional
style (M = 5.08), and subjects below the median were
lowing experience with uncontrollable noise, considered to have a specific attributional style (M = 2.87).
because the test situation is dissimilar to the Because the attributional dimensions for negative events
pretreatment. Subjects with a specific attri- are intercorrelated (Peterson et al., 1982), scores for in-
butional style for negative outcomes should ternality and stability were also computed for use as co-
show fewer deficits, or no deficits, on the test variates in subsequent analyses.
Subjects also completed the BDI (Beck, 1967), a self-
task in the dissimilar situation. report questionnaire that assesses the severity of depressive
The logic underlying these predictions is that symptoms. This instrument has an internal reliability in
all subjects who become helpless on the pre- the .80s (e.g., Peterson, Schwartz, & Seligman, 1981). It
treatment task, regardless of attributional style, was also included here for use as a covariate, because
attributional style is related to depression (Raps, Peterson,
also will be likely to expect to be helpless in Reinhard, Abramson & Seligman, 1982; Seligman et al.,
a very similar setting, because, presumably, 1979) and depression is associated with poor performance
the cause of their helplessness on the first task at tasks like those used here (Miller & Seligman, 1975).
will be operative in the similar situation. On Pretreatment apparatus. The pretreatment apparatus
is similar to that of Hiroto and Seligman (1975). Subjects
the other hand, subjects who attribute their sat in front of a small box with two lights and three buttons
helplessness on the first task to a specific cause on its top. They received 50 trials of signaled 90 db white
will not expect this cause to be operative in a noise through headphones. Each burst of noise lasted 5 s
dissimilar situation and therefore will not be unless terminated, with a mean intertrial interval of 14 s
(range - 10-25 s). Subjects in the controllable-noise group
likely to expect to be helpless on the task in could terminate the noise once it started by pushing one
the dissimilar setting. Subjects who attribute of the buttons. The lights indicated whether the noise hdd
their helplessness on the first task to a global been successfully terminated. Uncontrollable-noise subjects
cause will expect this cause to be operative in were yoked subject-by-subject to controllable-noise sub-
a wide variety of situations and therefore will jects, receiving the identical amounts of noise. The no-
noise subjects simply sat for a period of time equivalent
be more likely to expect to be helpless in a to the pretreatment prior to engaging in the test task.
dissimilar situation as well as a similar one. Test task. The "similar" test task of Experiment 1 was
the hand shuttlebo* of Hiroto and Seligman (1975). This
Experiment 1 device is a 24 X 5 X 6 in. (61 X 13 X 15 cm.) box with
a 3-in. (7 cm.) knob protruding from the top and sliding
Method on a 19-in. (48 cm.) straight channel. A warning light at
the middle of the box came on 5 s before and terminated
Subjects. Subjects were 108 undergraduates' at the when the 5-s 90 db noise delivered through headphones
University of Pennsylvania, recruited through notices began. Twenty trials were given (mean intertrial interval =
placed on campus bulletin boards, and paid $3 for their 21 s, range =10-45 s). The noise could be terminated
participation in the research. Subjects completed the ASQ by moving the knob to one side of the box after it started.
and the Beck Depression Inventory (BDI; Beck, 1967) On the next trial, the appropriate escape response was
approximately 1 week prior to participation in the lab- moving the knob to the opposite side of the box.
oratory experiment. Equal numbers of men and women Procedure. Approximately I week after completing
participated; sex of subject did not affect the pattern of the ASQ and BDI, subjects were individually tested in
results and is not mentioned further. single sessions conducted in the same room by the same
Questionnaires. The ASQ is described in detail by experimenter. For the pretreatment task, subjects were given
Peterson et al. (1982). Briefly, it provides six positive events these instructions:
(e.g., a friend compliments you on your appearance) and
six negative events (e.g., you go out on a date, and it goes From time to time, a loud tone will come on for a while.
badly). Respondents write down the major cause for each When that tone comes on, there is something you can
event if it happened to them and then rate each of their do to stop it.
chosen causes on 7-point scales in accord with its perceived
internality, stability, and globality (see Abramson et al.,
1978).
1
Means'across events are calculated for each of the three Experiment 1 used more subjects than Experiment 2
dimensions, separately for positive and negative events. (108 vs. 60), because we wished to be conservative and
Peterson et al. (1982) reported internal reliabilities for maximize the possibility of finding the nonpredicted in-
these dimensions, estimated by Cronbach's (1951) coef- teraction in Experiment 1.
684 ALLOY, PETERSON, ABRAMSON, AND SELIGMAN

There are two lights located on this base. The lights will Table 2
tell you how the tone on each trial was controlled. If Mean Latencies (in s) in the
the green light goes on following the termination of the Similar-Test Situation
tone, then your responses turned it off. If the red light
goes on following termination of the tone, then your Global Specific
responses did not turn it off, but rather it went off au- attributional attributional
tomatically according to a preprogrammed schedule. Pretreatment style style
Taking the earphones off or dismantling the apparatus
is not the way to stop the tone. Escapable noise
Immediately after the pretreatment, subjects were shown M 2.26, 1.98a
the shuttlebox, which had been covered to that point. SD 1.11 0.75
They were given these instructions: Inescapable noise
M 3.43b 3.43b
You will be given some trials in which a loud tone will SD 1.17 1.28
be presented to you. Whenever you hear the tone come No noise (control)
on, there is something you can do to stop it. Taking the M 2.54, 2.58a
earphones off or dismantling the apparatus is not the SD 1.25 1.46
way to stop the noise.
Note. Maximum mean latency = 5.00 s. For each condition,
n = 18. Means in a column or row are significantly different
Results (p < .05) by the Newman-Keuls test if they have different
subscripts.
A helplessness effect (i.e., debilitated per-
formance following experience with uncon-
trollable noise) was present for subjects with 5.06 on this dimension, whereas subjects with a specific
either a global or specific attributional style. attributional style averaged 2.87.
Average latencies2 were computed across trials Pretreatment apparatus. Experiment 2 used the iden-
tical pretreatment apparatus as Experiment 1.
on the test task for each subject (see Table 2) Test task. The "dissimilar" test task of Experiment 2
and analyzed in a 2 (attributional style) X 3 was the anagram test task of Hiroto and Seligman (1975):
(pretreatment) between-subjects analysis of 20 five-letter anagrams scrambled according to the same
variance (ANOVA). (If a burst of noise was not sequence (34251). Each anagram was printed on an index
card and was individually given to the subject, who had
terminated, a latency of 5 s was assigned.) up to 100 s to solve it. In addition, the test task was
A main effect of pretreatment was found, administered by a different experimenter in a different
F(2, 102) = 11.40, p < .001. Neither attri- room as part of a "second" experiment, in order to max-
butional style nor its interaction with the pre- imize the perceived dissimilarity between the pretreatment
treatment approached statistical significance and test situation (cf. Roth & Kubal, 1975).
Procedure. The same procedure as in Experiment 1
(.Fs < 1). Essentially the same main effect of was followed, except that the test task instructions were
pretreatment was found, F(2,99) = 11.15, p < changed:
.001, when scores from the BDI and the in- You will be asked to solve some anagrams. As you know,
ternality and stability dimensions of the ASQ anagrams are words with the letters scrambled. The
were used as covariates. As Table 2 shows, the problem for you is to unscramble the letters so that
pretreatment effect resulted frtim the poor they form a word. When you've found the word, tell
performance of uncontrollable-noise subjects me what it is.
relative to subjects in the other conditions
(Newman-Keuls comparisons among means Results
are presented in Table 2). A helplessness effect occurred for subjects
with a global attributional style for negative
Experiment 2
outcomes exposed to uncontrollable noise but
Method not for those with a specific attributional style
Subjects. Subjects were 60 undergraduates recruited
for negative outcomes. Average latencies (see
and paid as in Experiment 1. Subjects completed the ASQ Footnote 2) for anagram solutions were cal-
and BDI approximately 1 week prior to participation in
the laboratory experiment. Again, equal numbers of men
and women participated, and sex of subject did not affect
2
the pattern of results. For the sake of brevity, only latencies are reported.
Questionnaires. The ASQ and BDI are described in Analysis using number of failures yielded the same con-
Experiment 1. In Experiment 2, subjects considered to clusions, as would be expected, because latencies and fail-
have a global attributional style for negative events averaged ures were substantially correlated.
ATTRIBUTIONAL STYLE AND LEARNED HELPLESSNESS 685

Table 3 lable noise. In contrast, although learned


Mean Latencies (in s) in the helplessness deficits were induced in subjects
Dissimilar-Test Situation with a specific attributional style for negative
Global Specific
outcomes, these deficits only generalized to a
attributional attributional test situation that was similar to the original
Pretreatment style style situation in which they were exposed to un-
controllable events. Insofar as we varied a
Escapable noise number of elements (e.g., type of task, exper-
M 16.4, 17.4a
SD 11.9 6.9 imenter, room, and experiment) to maximize
Inescapable noise the perceived dissimilarity between the pre-
M 41.5b 20.0. treatment and test situations in Experiment
SD 26.8 11.5 2, future parametric studies are necessary to
No noise (control)
M 20.7, 19.0a
determine which of these manipulations alone
SD 13.6 12.2 or in combination was responsible for our ex-
perimental effects. In addition, although our
Note. Maximum mean latency = 100.0 s. For each con- results were consistent with the theoretical
dition, « = 10. Means in a column or row are significantly
different (p < .05) by the Newman-Keuls test if they have prediction that attributional style influences
different subscripts. the generality of learned helplessness, the ab-
solute amount of generalization exhibited by
subjects with a global attributional style in our
culated for each subject (see Table 3) and an- laboratory experiment was relatively narrow
alyzed in a 2 X 3 between-subjects ANOVA. (If compared with the widespread helplessness
an anagram was not solved, a latency of 100 deficits observed among, say, some concentra-
s was assigned.) tion camp inmates or severely depressed people
A main effect of pretreatment was found, in the natural environment (cf. Seligman,
F{2, 54) = 4.66, p < .05, but this is best in- 1975). Future work should address the rele-
terpreted in light of the significant interaction vance of the attributional style construct in
between pretreatment and attributional style, mediating more extreme cases of generaliza-
F(2, 54) = 3.30, p < .05. This interaction re- tion of learned helplessness deficits outside of
sulted from the poor performance by subjects the laboratory.
with a global attributional style for negative These results also add to the construct va-
outcomes following experience with uncon- lidity of the ASQ (Peterson et al., 1982). Re-
trollable noise (see Table 3 for Newman-Keuls sponses to the ASQ predicted which subjects
comparisons among means). Attributional would show relatively generalized problem-
style per se did not have a significant effect. solving deficits following experience with un-
The same interaction term, F(2, 51) = 3.97, controllability and which subjects would not.
p < .05, was obtained when scores from the According to the reformulated helplessness
BDI and the internality and stability dimen- theory, the reason subjects with a global at-
sions of the ASQ were used as covariates. tributional style show more generalized per-
formance deficits is that they attribute their
General Discussion helplessness to a global cause that is likely to
influence performance in situations both sim-
The results of the study were consistent with ilar and dissimilar to the original situation in
the prediction of the reformulated model of which they became helpless. An important is-
learned helplessness (Abramson et al., 1978) sue is whether subjects spontaneously make
that the individual-difference variable of at- causal attributions. The fact that subjects will
tributional style is a determinant of the gen- rate the causes of hypothetical outcomes on
erality of learned helplessness. Subjects who the ASQ does not necessarily mean that they
exhibited a global attributional style for neg- will use such attributions in the absence of
ative outcomes (as measured by the ASQ) specific prompting to do so. Recently, several
showed learned helplessness deficits that gen- investigators (Enzle & Schopflocher, 1978;
eralized to both a similar and a dissimilar sit- Pittman & Pittman, 1980; Pyszczynski &
uation following experience with uncontrol- Greenberg, 1981; Swann, Stephenson, & Pitt-
686 ALLOY, PETERSON, ABRAMSON, AND SELIGMAN

man, 1981; Wong & Weiner, 1981) have begun their helplessness, should only expect this cause
to specify conditions leading to the initiation to be present in the similar test situation. Ac-
of attributional processing. Two main deter- cording to Abramson et al. (1978), if the ex-
minants of the onset of spontaneous causal pectation of uncontrollability is present in a
analyses have been identified in this research: given situation, helplessness behavioral deficits
experience of failure or loss of control (Pittman will ensue. Thus subjects with a global attri-
& Pittman, 1980; Swann et al., 1981; Wong butional style should exhibit poor performance
& Weiner, 1981) and unexpected behaviors or in both the similar and dissimilar test situa-
events (Pyszczynski & Greenberg, 1981; Wong tions, whereas subjects with a specific attri-
& Weiner, 1981). Both of these factors were butional style should exhibit poor performance
likely to be operative in the present experi- only in the similar test situation. This is, of
ments; subjects exposed to uncontrollable course, the pattern of results we obtained. To
noises experienced loss of control and unex- test the validity of the mediational processes
pected failure in problem solving. Thus, it is hypothesized by Abramson et al., it will be
likely that subjects in the uncontrollable-noise necessary to assess the intervening causal at-
conditions actually made attributions for their tributions for the particular uncontrollable
helplessness. event as well as subsequent expectations of
An important task for future research is the helplessness (Alloy, 1982).
specification of the psychological processes that Because our study uses a design that was
mediate the relation between attributional style partially correlational, it is possible that con-
and the generality of learned helplessness. We trary to the reformulated helplessness theory,
emphasize that our study was designed to de- some third variable mediated the relation be-
termine whether such a relation exists but not tween globality of attributional style and gen-
to examine the processes mediating the rela- erality of learned helplessness. Given the pre-
tion. According to the reformulated theory of viously reported positive correlation between
helplessness (Abramson et al., 1978), the se- individuals' level of depression and their ten-
quence of mediating events is as follows: At- dency to make global attributions for negative
tributional style influences the causal attri- outcomes (e,g., Seligman et al., 1979), one
bution subjects make for a particular uncon- possible mediating third variable is subjects'
trollable event, and this causal attribution for initial level of depression. However, the fact
the particular uncontrollable event influences that our findings remained unchanged when
subsequent expectations of helplessness, subjects' initial BDI scores were used as co-
which, in turn, determine the generality of variates in the statistical analyses is not con-
learned helplessness: Applying the logic of the sistent with this particular third variable hy-
mediational process postulated by the refor- pothesis. Alternative potential third variables
mulated helplessness theory to our particular include subjects' tendencies to make internal
experimental setting, subjects who exhibited or stable attributions for negative outcomes.
a global attributional style for negative out- However, these third variables also did not
comes should have made a more global attri- appear to be operative because the pattern of
bution for their inability to control noise than findings remained unchanged when subjects'
subjects who exhibited a specific attributional scores along these attributional dimensions
style (e.g., to general stupidity vs. inability at were used as covariates. An alternative third
mechanical tasks). Because a global attribution variable hypothesis, which could not be ex-
for the cause of helplessness has implications amined in this study, is that subjects who ex-
for performance in a wider variety of settings hibit global attributional styles for negative
than does a specific attribution for helplessness, events experience more depressive affect fol-
subjects with a global attributional style should lowing experience with an unsolvable problem
expect this cause to be present and thus expect (cf. Metalsky, Abramson, Seligman, Semmel,
to be helpless in the dissimilar test situation & Peterson, 1982) and that this depressive af-
(anagrams, Experiment 2) as well as the similar fect rather than their particular causal attri-
test situation (shuttlebox, Experiment 1). Sub- butions and expectations mediates the relation
jects with a specific attributional style, who between global attributional style and gener-
presumably make a specific attribution for alized learned helplessness. Thus, future stud-
ATTRIBUTIONS, STYLE AND LEARNED HELPLESSNESS 687

ies using experimental manipulations and attribution processes by attributional questions. Per-
sonality and Social Psychology Bulletin, 4, 595-599.
causal modeling techniques will be required Hiroto, D. S., & Seligman, M. E. P. (1975). Generality of
to pit the mediational hypothesis proposed by learned helplessness in man. Journal of Personality and
the reformulated helplessness theory against Social Psychology, 31, 311-327.
the depressive affect mediational hypothesis as Maier, S. E, & Seligman, M. E. P. (1976). Learned help-
well as other rival mediational hypotheses. lessness: Theory and evidence. Journal of Experimental
Psychology: General, 105, 3-46.
In their reformulated model of learned Metalsky, G. I., Abramson, L. Y., Seligman, M. E. P.,
helplessness and depression, Abramson et al. Semmel, A., & Peterson, C. (1982). Attributional styles
(1978) hypothesized that certain attributional and life events in the classroom: Vulnerability and in-
styles are vulnerability factors for depressive vulnerability to depressive mood reactions. Journal of
reactions to negative life events. Our results Personality and Social Psychology, 43, 612-617.
Miller, I. W., & Norman, W. H. (1979). Learned help-
show that people who exhibit a style of at- lessness in humans: A review and attribution theory
tributing negative outcomes to global causes model. Psychological Bulletin, 86, 93-119.
show greater generalization of learned help- Miller, W. R., & Seligman, M. E. P. (1975). Depression
lessness to new situations than do people who and learned helplessness in man, Journal of Abnormal
Psychology, 84, 228-238.
do not exhibit the global attributional style Pasahow, R. J., West, S. G., & Boroto, D. R. (1982). Pre-
for negative outcomes. Complementing these dicting when uncontrollability will produce performance
findings, Metalsky et al. (1982) recently con- deficits: A refinement of the reformulated learned help-
ducted a study in a naturalistic setting and lessness hypothesis. Psychological Review, 89, 595-598.
found that the more global a student's attri- Peterson, C, Schwartz, S. M., & Seligman, M. E. P. (1981).
Self-blame and depressive symptoms. Journal of Per-
butional style for negative outcomes, the more sonality and Social Psychology, 41, 253-259,
severe was his or her depressive mood reaction Peterson, C., Semmel, A., von Baeyer, C., Abramson,
to a negative life event (i.e., receipt of a low L. Y., Metalsky, G. I., & Seligman, M. E. P. (1982).
midterm grade; r = .32). Taken together, these The attributional style questionnaire. Cognitive Therapy
and Research, 6, 287-299.
studies suggest that people with global attri- Pittman, T. S., & Pittman, N. L. (1980). Deprivation of
butional styles for negative outcomes are par- control and the attribution process. Journal of Person-
ticularly vulnerable to debilitating behavioral ality and Social Psychology, 39, 377-389.
and emotional responses to negative events. Pyszczynski, T. A., & Greenberg, J. (1981). Role of dis-
continued expectancies in the instigation of attributional
Because both studies used prospective designs, processing. Journal of Personality and Social Psychology,
they may have implications for identifying in- 40, 31-38.
dividuals who are at high risk for "helplessness Raps, C. S., Peterson, C., Reinhard, K. E., Abramson,
depression" (Seligman, 1978) and intervening L. Y., & Seligman, M. E. P. (1982). Attributional style
with strategies designed to help these individ- among depressed patients. Journal of Abnormal Psy-
chology, 91, 102-108.
uals modify their deleterious attributional Roth, S., & Kubal, L. (1975). Effects of noncontingent
styles. reinforcement on tasks of differing importance: Facil-
itation and learned helplessness. Journal of Personality
and Social Psychology, 32, 680-691.
References Seligman, M. E. P. (1975). Helplessness: On depression,
development, and death. San Francisco: W. H. Freeman.
Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. Seligman, M, E. P. (1978). Comment and integration.
(1978). Learned helplessness in humans: Critique and Journal of Abnormal Psychology, 87, 165-179.
reformulation. Journal of Abnormal Psychology, 87,49- Seligman, M. E. P., Abramson, L. Y., Semmel, A,, & von
74. Baeyer, C. (1979). Depressive attributional style. Journal
Alloy, L. B. (1982). The role of perceptions and attributions of Abnormal Psychology^, 88, 242-247,
for response-outcome noncontingency in learned help- Swann, W.B., Jr., Stephenson, B., & Pittman, T. S. (1981).
lessness: A commentary and discussion. Journal of Per- Curiosity and control: On the determinants of the search
sonality, 50, 443-479. for social knowledge. Journal of Personality and Social
Beck, A. T. (1967). Depression: Clinical, experimental, Psychology, 40. 635-642.
and theoretical aspects. New York: Harper & Row. Wong, P. T. P., & Weiner, B, (1981). When people ask
Cole, C. S., & Coyne, J. C. (1977). Situatkmal specificity "why" questions, and the heuristics of attributional
of laboratory-induced learned helplessness in humans, search. Journal of Personality and Social Psychology,
Journal of Abnormal Psychology, 86, 615-623. 40, 650-663.
Cronbach, L. J. (1951). Coefficient alpha and the internal
structure of tests, fsychometrika, 16, 297-334. Received August 9, 1982
Enzle, M. E., & Schopflocher, D. (1978). Instigation of Revision received January 26, 1983 •

You might also like