Self-handicapping following learned helplessness treatment and the type A coronary-prone behavior pattern

Self-handicapping following learned helplessness treatment and the type A coronary-prone behavior pattern

Journal of P\ychowmaric Rcrcarch. Vol. 24. pp. 3lY-325 Pergamon Prei\ 1 Id. 1980. Pnnled in Great Bnlain. SELF-HANDICAPPING HELPLESSNESS FOLLOWING ...

590KB Sizes 106 Downloads 71 Views

Journal of P\ychowmaric Rcrcarch. Vol. 24. pp. 3lY-325 Pergamon Prei\ 1 Id. 1980. Pnnled in Great Bnlain.

SELF-HANDICAPPING HELPLESSNESS

FOLLOWING

TREATMENT

CORONARY-PRONE

(Received

2 February

AND THE TYPE A

BEHAVIOR

GERDI

LEARNED PATTERN

WEIDNER

1980; in revisedform

9 June 1980)

Abstract-This investigation examined the choice of either a performance enhancing or a performance inhibiting drug among coronary-prone (Type A) and non-coronary-prone (Type B) men after contingent feedback (success) or non-contingent feedback (failure) on a task. All subjects expected to work on a similar task under the influence of the drug. Type A’s in the failure condition chose the performance inhibiting drug significantly more often than Type A’s in the success condition or Type B’s in either condition. The results were discussed in terms of differences in attributional strategies among Type A’s and B’s, which could possibly mediate the relationship between Type A, helplessness and heart disease. EVIDENCE linking life stress with coronary heart disease (CHD) has accumulated during the past few years. For example, Rahe and Lind [I] demonstrate that sudden cardiac death is often preceded by an increase in life stress during the 6-month period before death. Greene et a/. [2] also observed that victims of CHD have been depressed for a week up to several months prior to death, possibly resulting from failure to cope with unexpected stressful life events. According to Seligman [3] there are major parallels between symptoms of depression and the state of helplessness. Learned helplessness occurs when an organism learns that its reinforcements are independent of its responses (i.e. lacking control over the environment), and this learning leads to giving up further instrumental responses on subsequent controllable events. Thus, it seems to be reasonable to assume that the state of helplessness may precede CHD. The level of stress a person perceives is determined not only by environmental factors but also by characteristics of the individual. Rosenman and Friedman and their colleagues [4, 51 have identified a coronary-prone behavior pattern, called Type A. Persons exhibiting this behavior pattern are characterized by competitive achievement striving, time urgency and aggression. Several retro- as well as prospective studies have shown that Type A’s are at least twice as likely to develop heart disease as persons with the opposing Type B behavior pattern [6,7]. There is some evidence that it is the person exhibiting the coronary-prone behavior pattern Type A who is particularly prone to experience helplessness [S]. The basic paradigm used for demonstrating helplessness in Type A’s is based on a modification of a procedure used in learned helplessness research [9]. Krantz et al. [8] have shown that when confronted with salient uncontrollable events, Type A’s increase their initial efforts to assert control. However, extended exposure to uncontrollability leads to greater helplessness in Type A’s than Type B’s. The precise reason for Type A’s giving up attempts to assert control when confronted with salient uncontrollable events is not clear. In a recent reformulation of the learned helplessness hypothesis, Abramson, Seligman and Teasdale [lo] Gerdi Weidner,

Department

of Psychology,

Kansas State University, 319

Manhattan,

KS 66506, U.S.A

320

C;EKL>I WEII)NPR

stated that once people perceive noncontingency between their responses and outcomes, they attribute their helplessness to a cause. The attributions an individual makes are seen as determinants of subsequent expectations for future noncontingency. These expectations will, in turn, determine both the kind of helplessness deficit and its generality and its chronicity. For example, Tennen and Eller [I l] demonstrated that subjects failing on problems labelled as “easy” (thus producing more internal attributions for failure, such as lack of ability) exhibited more signs of learned helplessness than subjects working on the same problems labelled as “difficult”. Attributions for failure to internal factors have not only been associated with performance deficit, but also with specific affective reactions. Abramson [12] has demonstrated that only attributions to internal factors are associated with lowered self-esteem. Since Type A’s show more pronounced signs of learned helplessness than Type B’s when faced with salient uncontrollable events, it is suggested that Type A’s and B’s may differ in terms of attributions made for this perceived noncontingency, i.e. Type A’s may make more internal attributions for failure than Type B’s and also experience loss of self-esteem. Recently, Berglas and Jones [13] have suggested that individuals eupecting to engage in a behavior that will indicate inferior competence, are likely to perform actions that enhance the opportunity to externalize failure. For example, the student expecting to do poorly on an exam might go out and party the night before in order to have an excuse for not performing well. Berglas and Jones have referred to the,e actions as self-handicapping strategies. In a laboratory demonstration of this phenomenon subjects were asked to choose either a drug that would inhibit or one that would enhance performance on a prospective task. Those who previously had received successful feedback that appeared to be contingent on appropriate knowledge chose the performance enhancing drug; those receiving a predetermined feedback (noncontingent on performance) chose the inhibiting drug. Herglas and Jones reasoned that the noncontingent success group wished IO externalize probable failure on the retest by attributing their prospective poor performance to the inhibiting effects of the drug, and thus protect their self-esteem. Since it was argued that Type A’s are more threatened by failure on a task than Type B’s, it might be expected that Type A’s in particular are prone to accept the opportunity to externalize prospective failure on a task. In sum, the present study was designed to investigate whether Type A’\ make more internal attributions for failure than Type B’s. Since internal attributiori~ for failure are associated with lowered self-esteem, it was also anticipated tka: fypc A’s would be more likely to engage in a self-handicapping strategy than T![?t. :!‘h, thus allowing them to externalize failure and preserve their self-esteem.

METHOD Overview and design To test the above hypotheses, Type A and Type B male college students were asked to choose between a performance enhancing drug, a performance inhibiting drug, or no drug afrer experiencing either success or failure on a concept formation task. All subjects believed that they had to complete a task similar to the previous one under the influence of the drug doses. The design was a 2 x 2 factorial with IO subjects per group. The independent variables were Type A, Type B and the evpewncc of failure or success on the concept formation task.

Self-handicapping

following

learned helplessness

321

treatment

Subjects The subjects of the study were 42 male undergradute students enrolled in introductory psychology at a large midwestern university. One subject declined to participate af!er he was informed of the purpose of the study* and another subject was excluded because he was suspicious about the procedure. Classification of the remaining 40 subjects as Type A’s or Type B’s was based on scores on the student version of the Jenkins Activity Survey (JAS) [14]. The median score was used to divide subjects into Type A’s (8.0 and above) and Type B’s (7.0 and below). Other than the requirement that each experimental condition contained equal numbers of Type A’s and B’s, subjects were randomly assigned to groups. Procedure The procedure used in this experiment was similar to the one employed by Berglas and Jones [13]. The experiment consisted of 2 sessions. In the first session subjects were run in small groups (335) by a female experimenter. An information sheet was provided explaining that the experiment had been designed to test whether either of two drugs used in the treatment of metabolic disorders had an effect on intellectual performance. Subjects were told that during the first session they would be filling out questionnaires concerning their personality characteristics (JAS) and a questionnaire to assess their medical history.? The second session, for which they could sign up later, would consist of the actual experiment. Subjects were told they would perform two similar tasks. After completing the first task they could choose one of two conditions under which they would take the second task: no drug, or either of the two drugs. Each subject was reminded of his opportunity to withdraw from the experiment at any time. The second session of the experiment was conducted by two female experimenters. Subjects arrived individually, were greeted by the first experimenter, who explained the purpose cf the study again, adding that the subject would be given detailed information about the drugs after completing the first task. Then the experimenter described the nature of the task. Subjects were told they would be working on four concept formation problems which consisted of a series of four-dimensional stimulus patterns [IS]. Each of the four dimensions had 2 associated values: (1) letter (A or T); (2) letter size (large or small); (3) color of the letter (red or blue); (4) border shape (square or circle). The subject’s task was to search for the one dimension that was correct. After a sample set of problems was used to familiarise the subject with the procedure, the four problems consisting of 10 stimulus cards (trials) were administered. Subjects in the success condition were given contingent reinforcement after each trial, thereby allowing them to deduce the correct value before the end of a 10 trial set. To be included in the data analysis, subjects had to get at least 3 out of 4 problems correct. In the noncontingent or failure condition, no value was treated as consistently correct. The subject was given a predetermined sequence of “correct and incorrect” answers. Also, subjects in the failure condition were told “That’s the wrong answer”, when they were asked to identify the correct value at the end of each dimensional problem. In addition, a tally sheet was handed to all subjects who were asked to keep a record of correct and incorrect answers. This procedure has been used by Glass [14] in the past to increase the salience of failure or success experiences. After the subject had completed the problems, the experimenter escorted him to another testing room and introduced him to the third experimenter, who was blind to the subject’s experimental conditions and his AB-score. The subject was then given detailed information about the drugs. Preliminary reports “from the Physician’s Desk Reference” for both Pandocrin and Actavil (actually bogus reports), explaining chemical composition, indication, dosage and precautions, were handed to the subjects. These reports included the benign side effects of each drug and the critical information about each drug’s effect on intellectual functioning. The report on Actavil noted that performance should be facilitated through “stimulation of associative processes and general heightening of cognitive acuity”. The report on Pandocrin described the performance inhibiting effect in terms of “decrements in cognitive association, attention, and powers of concentration”. The subject was presented with a graph illustrating the results of a bogus study from “Journal of Psychopharmacology, in press”. The graphs showed improvement on intellectual performance for subjects who had taken Actavil; deteriorated performance effects were depicted for Pandocrin. It was added that these results were obtained from a sample of middle-aged adults, and it was not certain whether the same effects would occur in younger people as well. Thus, the present experiment was necessary in order to gather more reliable data. Following a brief “medical” examination consisting of blood pressure measurement and pupillary dilation check, the subject was told that a physician had checked his medical history which he had filled out during the first session and had okayed his participation. Furthermore, the subject was assured that Actavil was not like an amphetamine nor did Pandocrin have the effect of a “downer”, and that the performance effects probably would occur without his awareness. Also, the subject was: informed that the *The title of the experiment was kept ambiguous to avoid selection of subjects drugs. are available from the author. *Copies of the “Medical History” questionnaire

interested

in taking

322

GERDI WEIDNER

effect of the drugs would be transitory. Then the subject was shown a tray on which the drugs were placed. The subject was instructed to choose 1 of 4 levels of either drug, the maximum dosage being 10 mg, followed by 7.5, 5.0 and 2.5 mg. The subject was told that there had been approximately the same number of people for each level of the drugs and the no drug condition. Thus, no pressure was put on the subject to take either drug. In addition, the experimenter stated: “I don’t want you to take either drug at any particular dosage level to please me. Select either drug according to what you will find most interesting.” After indicating the drug and which dosage level he decided to take, the subject was asked to complete a series of manipulation checks entitled “Mid-task questionnaire”. Questions concerning difficulty of the task, attributions of task performance (luck vs ability), and assessment of effort put into solving the problems were included. In addition, subjects were asked to indicate how they thought the chosen drug would affect certain aspects of their performance, including number of correct trials and number of correct final guesses. The experiment was then terminated. No drug was given to the subjects and no second set of problems was administered. The subject was fully debriefed and thanked for participation.

RESULTS

The postexperimental questionnaires and the drug choice were analyzed by a series of 2 x 2 analyses of variance. The independent variables were Type A-B and failure-success condition.

Manipulation checks In order to confirm that the success-failure treatments were perceived differentially, subjects were asked to indicate on 7-point scales how difficult they found the task and how well they thought they had solved the problem. Subjects in the failure condition rated the task more difficult than subjects in the success condition, F (1, 36) = 61.6, P < 0.001. also, the problem was believed to be better solved by subjects in the success than by subjects in the failure condition, F (I, 36) = 110.6, P < 0.001. No other significant interactions or main effects were found. Thus, it was concluded that the manipulation of induced experiences of failure and success was effective. In order to check on subjects’ belief in the effectiveness of the two drugs, their “Given the choice of drug that ratings on the following questions were obtained: you made-how well do you think you will do (relative to the first task) on the following aspects of performance: number of correct final guesses, number of correct trials”. Ratings were made on 7-point scales where 1 equalled “definite11 less” and 7 equalled “definitely more”. To facilitate an overall statistical analysis, drug dosage and drug choice were assigned values on a single 9-point scale. Scale values for the different dosage levels of the drug were assigned running from 1 = highest dosage of the inhibiting drug to 9 = highest dosage of the enhancing drug. We would expect lower scale values (reflecting the inhibiting drug) to be associated with lowered expectations of doing better. To test this prediction, Pearson Product Moment Correlation coefficients were calculated, revealing that drug scale value and number of correct final guesses were related (r = 0.40, P < 0.02). Drug scale values correlated with number of correct trials did not achieve the desired level of significance (r = 0.24, P < 0.12). A significant correlation was obtained between drug scale values and subjects’ rated beliefs that they would do better or worse in general on the next task under the influence of the drug (r = 0.42, P < 0.01). It was thus concluded that subjects believed that their chosen drug and chosen drug level would have the suggested effect on further task performance.

Self-handicapping

following

learned

helplessness

treatment

323

Drug choice It was hypothesized that subjects in the failure condition would be more prone to self-handicap themselves (i.e. to choose the performance inhibiting drug) than would subjects in the success condition. There was a significant main effect for success-failure condition using the scale value measures of drug choice; F (1, 36) = 14.7, P < 0.001, indicating that subjects in the success condition chose higher dosage values (indicative of the performance enhancing drug) than subjects in the failure condition. Thus, more self-handicapping was found in the failure than in the success condition, paralleling Berglas and Jones’ [13] finding on the use of self-handicapping strategies after noncontingent feedback. However, this effect was attributable largely to Type A’s drug choice in the failure condition as suggested by a significant interaction, F (1, 36) = 4.6, P < 0.05. Planned comparisons revealed that Type A’s in the failure condition chose lower scale values (indicative of the performance inhibiting drug) than did Type A’s in the success condition or Type B’s in either condition (P’s < 0.05). Table 1 summarizes the effects of Type and condition on drug choice. TABLE

I.-EFFECTS

CONDITION

ON

THE

OF

A-B

TYPE

CHOICE

OF THE

AND

PERFORMANCE

SUCCESS/FAILURE ENHANCING/

INHIBITINGDRUGS

Condition Success Failure

A

B

7.8 4.6

7.3 6.4

Scale values run from 1, the highest dosage of the inhibiting drug (Pandocrin), to 9, the highest dosage of the enhancing drug (Actavil).

Thus, the hypothesis that people in the failure condition would choose more to self-handicap themselves than people in the success condition was confirmed due to Type A’s relative preference for the performance inhibiting drug.

Type A-B differences in reported attributions In order to assess whether self-handicapping would be mediated by differential attributions, a number of post hoc analyses were performed. Comparisons of attributions of Type A’s in the failure condition with Type A’s in the success condition revealed a marginally significant finding. Type A’s in the failure condition tended to report less effort than Type A’s in the success condition (Newman-Keuls, P < O.lO), whereas Type B’s did not differ in either condition. Thus, there seems to be only slight evidence for differences in reported attributional strategies (effort) among Type A’s experiencing either success or failure.

DISCUSSION

These findings confirm the prediction that men faced with the prospect of failing at a task are more likely to adopt a self-handicapping strategy than men expecting to succeed. This pattern of results, however, was due to Type A’s drug choice in the

324

GLKDIWEIDNEK

failure condition. Type B’s did not choose to self-handicap themselves in either condition. According to Berglas and Jones [13], the choice of self-handicapping in view of prospective failure provides an opportunity to externalize or excuse bad performance, and thus protects one’s self-esteem. The finding that Type A’s are likely to self-handicap themselves when faced with failing at a task has some important theoretical implications. According to Abramson et a/. [lo], self-esteem deficits follow from attributions of failure to internal factors that are controllable (e.g. lack of effort). Some evidence was found in the present study that Type A’s indeed attributed their failure to internal factors. Type A’s in the failure condition tended to report less effort invested into solving the problems than Type A’s in the success condition, whereas Type B’s did not differ across conditions. It is possible that the measures were not sensitive enough to detect differences in attributional strategies between the groups. In a recent experiment Brunson [ 161 found consistency between behavioral and cognitive measures, requiring subjects to continuously verbalize during task failure. Using this phenomenological approach, he found that Type A’s in a failure condition made more internal attributions than did Type B’s. Thus, a continuous verbalization procedure may be more sensitive in detecting differential experiences of failure than a simple verbal self-report as employed in the present study. Regardless of the underlying mechanisms mediating self-handicapping behavior, the finding that Type A’s were likely to take the opportunity to excuse bad performance by attributing failure to an external source (i.e. choice of a performance inhibiting drug) is significant. According to Monat and Lazarus [17], stress may lead people to engage in coping strategies that are damaging to their health. The observation that helplessness inducing life events often precede heart disease [2], together with the empirical finding that Type A’s are particularly prone to experience helplessness when confronted with stressful uncontrollable events [8], suggests that Type A’s coping strategy with stressful events may be maladaptive in the long run. The present finding that Type A’s were likely to self-handicap themselves in the view of prospective failure contributes to the understanding of Type A’s coping strategies when confronted with salient uncontrollable events. Type A’s may not only passively give up attempting to assert control after exposure to noncontingent events, but they may actively engage in behavior that prevents them from experiencing response-outcome dependency and success. This finding has some practical application. To eliminate learned helplessness effects, the organisms are usua11~~ exposed to situations in which they learn that their responses do have an impact on the environment [9]. However, mere exposure to contingency of responses and outcomes may not be an effective intervention technique for those Type A’s who actively engage in behaviours that prevent them from experiencing control over the environment. Since there is some evidence that Type A’s may make internal attributions for failure associated with lowered self-esteem, it is reasonable to assume that attributions mediate the relationship between Type A behavior, helplessness, and coronary heart disease. Thus, focusing on the attributional process directly as well as teaching Type A’s to make accurate assessments of their ability to control their environment may result in an improved coping with unexpected stressful events.

Self-handicapping Acknowle&ements-The author comments on this manuscript.

thanks

following

learned helplessness

Brad Brunson,

William

Griffitt

treatment

325

and Joe Istvan for their helpful

REFERENCES 1. RZHF R. H. and LIND E. Psychosocial factors and sudden cardiac death: A pilot study. J. Psychosom. Rex 15, 17 (1971). aspects of sudden death: a preliminary 2. GREENE W. A., GOLDSTEIN S. and Moss A. J. Psychosocial report. Archs Intern. Med. 124, 725 (1972). and learned helplessness. In The Psychology of Depression: 3. SELIGMAN M. E. P. Depression Confemporary Theory andResearch (Edited by FRIEDMAN R. J. and KATZ M. M.). Winston-Wiley, Washington (1974). 4. ROSENMANR. H., FRIEDMAN M., STRAUSSR., WURM M., KOSITCHEI( R., HAHN W. and WERTHESSFN N. T. A predictive study of coronary heart disease. J. Am. Med. Ass. 189,103 (1964). 5. ROSENMAN R. H., FRIEDMAN M., STRAUSS R., WURM M., JENKINS C. D. and MESSINGER H. B. Coronary heart disease in the Western Collaborative Group study: a follow-up experience of two years. J. Am. Med. Ass. 195, 130 (1966). 6. ROSENMAN R. H., BRAND R. J., JENKINS C. D., FRIEDMAN M., STRAUSS R. and WURM M. Coronary disease in the Western Collaborative Group study: final follow-up experience of 8 % years. J. Am. Med. Ass. 233, 872 (1975). 7. KENICSBERG D., ZYZANSKI S. J., JENKINS D. D., WARDWELL W. T. and LICCIARDEL~O A. T. The coronary-prone behavior pattern in hospitalized patients with and without coronary heart disease. Psychosom. Med. 36,344 (1974). stress level, and the coronary-prone 8. KRANTI. D. S., GLASS D. C. and SNYDER M. L. Helplessness, behavior pattern. J. ,%p. Sot. Psycho!. 10,284 (1974). 9. SELXMAN M. E. P. Helplessness: On Depression, Development and Death. Freeman, San Francisco (1975). 10. ABRAMSON L. Y., SELIGMAN M. E. P. and TEASDALE J. D. Learned helplessness in humans: critique and reformulation. J. Abnorm. Psycho/. 87,49 (1978). 11. TENNEN H. and ELLER S. J. Attributional components of learned helplessness and facilitation. .I. Personality Sot. Psychol. 35,265 (1977). test of the reformulated 12. ABRAMSON L. Y. Universal versus personal helplessness: an experimental theory of learned helplessness and depression. Unpublished doctoral dissertation, University of Pennsylvania (1977). strategy in response to non13. BERCLAS S. and JONES E. E. Drug choice as a self-handicapping contingent success. J. PersonalitySot. Psycho/. 36,405 (1978). Stress, and Coronary Disease. Lawrence Erlbaum associates, 14. GLASS D. C. Behavior Patterns, Hillsdale (1977). 15. LEVINE M. Hypothesis behavior by humans during discrimination learning. J. Exp. Psychol. 71, 331 (1966). behavior pattern and reactions to uncontrollable events: 16. BRUNSON B. I. The Type A coronary-prone an analysis of learned helplessness. Unpublished Master’s thesis, Kansas State University (1979). 17. MONAD A. and LAZARUS R. S. Stress and Coping. Columbia University Press, New York (1977).