_ FEAR OF FAILURE AND ROLE . CONGRUENCE: AN INVESTIGATION ~ INTO THE NATURE OF ACHIEVEMENT , ' MOTIVATION IN WOMEN Dussertatlon for the Degree 0th D 5 I MICHIGAN STATE UNIVERSITY ‘ “ ALICE MADELINE WOLFE BERNSTEIN 51:12.5: T _, 1975 1‘. III!IIIIILIIIIIIIIIIIIIIIIIIIIIJIIILIIII I This is to certify that the " thesis entitled Fear of Failure and Role Congruence: An Investigation into the Nature F I‘ of Achievement Motivation in Women presented by Alice Madeline Wolfe Bernstein has been accepted towards fulfillment ‘ of the requirements for i * ‘ . Ph.D. (“agree in Psychology ' Major professor . e . . _ 1 Dnte/%7j.:/S 3/;yf/ ' . k . L ‘».- / - ’ ‘ @! A«...f. 5’ .3 ’1‘ ‘1 l." 4‘» 4 ‘ O~7639 I' ._ Ll VII II ,,/ fle‘l'fiiah ' ' . N *—W24 ZINE—fl ABSTRACT FEAR OF FAILURE AND ROLE CONGRUENCE: AN INVESTIGATION INTO THE NATURE OF ACHIEVEMENT MOTIVATION IN WOMEN BY Alice Madeline Wolfe Bernstein This study was designed to demonstrate that the achievement behavior of women can be explained by the expectancy—value theory of achievement motivation without recourse to the concept of motive to avoid success. Pre- dictions based upon the inhibitory model of motive to avoid failure and on previous studies relating ego involve- ment to performance were made about performance, level of aspiration, and level of expectation. On the basis of their answers to three questions, female volunteers from introductory psychology classes were classified as having a career orientation, a home- making orientation, or as being undecided about their sex role orientation. The undecided group was eliminated. Stories told to four relatively neutral sentences were Scored for hostile press imagery. A median break classified subjects with any hostile press imagery as high fear of failure subjects. At a second session the 100 remaining subjects, divided into four groups on the Alice Madeline Wolfe Bernstein basis of their sex role orientation and fear of failure classification, were administered a scrambled words test. Half of the subjects within each group were told that the task was a measure of homemaking ability, and half were told that the task was a measure of career ability. Before attempting the task the subjects were asked to indicate their levels of aspiration and expectation. The level of aspiration, level of expectation, performance, and post—performance estimates of career- oriented subjects were significantly higher than those of subjects with a homemaking orientation. The task con- gruence variable produced significant F ratios for per- formance and numerical estimate of performance. Although the fear of failure subjects had an overall nonsignificant tendency to perform better than subjects without fear of failure, their post-performance percentile estimates had a significant tendency to be lower than those of subjects without fear of failure. 1 When the task was described as a test of homemaking ability, subjects with fear of failure performed better than subjects without fear of failure (p < .051). Career—oriented subjects with fear of failure set their level of aspiration high and appeared to under- estimate their own performance while over—estimating the While they estimated that their performance of others. ability to succeed at a career was well above the average, Alice Madeline Wolfe Bernstein they estimated that their ability to be a homemaker was below average. Fear of failure subjects with a homemaking orientation, who were significantly more numerous than those with a career orientation, set their level of aspiration extremely low and estimated that their per- formance was extremely low. Relative to the other sub- jects, their estimates of ability on tests of career and homemaking ability were low, but there was little dif- ference between their two ability predictions. It is hypothesized that incentive value has both a competence and a social consequences component and that the relative weight of the social consequences component is greater for the fear of failure individual than for the individual without fear of failure. Fear of failure is conceived of as a strategy designed to minimize social rejection while maintaining a sense of personal compe- tency. Approved Elaine Donelson Committee Chairperson Jean Gullahorn Donald Grummon Barbara Riemer Committee Members Date FEAR OF FAILURE AND ROLE CONGRUENCE: AN INVESTIGATION INTO THE NATURE OF ACHIEVEMENT MOTIVATION IN WOMEN BY Alice Madeline Wolfe Bernstein A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY I Department of Psychology 1975 To Larry Who washed the dishes while I wrote. ii ACKNOWLEDGMENTS Without the help of many people it would have been impossible to finish my dissertation in East Lansing while living in Chicago. First, and foremost, I would like to thank Dr. Elaine Donelson, my Committee Chair- ‘ person, for her inspiration, guidance, and her toleration of the many long delays and gaps in communication caused by my physical separation from the campus. I would also like to thank the other members of my committee, Dr. Jean Gullahorn, Dr. Don Grummon, Dr. Gail Zellman, and Dr. Barbara Riemer, for their contributions. The experiment was conducted by Kathy Miller and Beth Oleshansky; their assistance was Vital and is greatly appreciated. Ed Bourne and Ellen Washington of the Uni— versity of Chicago helped me with the reliability checks. I Only hOpe that I am as good a friend to them as they are to me. To Roger Ahrends I extend my thanks for helping me Score the scrambled words protocols. Sue Wessner of the Graduate Research Office deserves great thanks for typing / the rough draft of this lengthy manuscript. The many variables examined in this study neces— sitated a computer analysis. Without Dave Thissen's help, I would have had to remain in East Lansing while analyzing the data. Sandra Stuart generously provided me to stay during my visits to the campus. She nave a warm place in my heart and a warm bed when she visits Chicago and wherever we move Rita Larson provided me with a place to stay with a place will always to stay in afterwards. during the early part of my work and also handed out the final draft :opies. Like Sandy, she will always be welcome in my home. This study was partially supported by a Graduate foice Fellowship awarded by the Department of Psychology, dichigan State University. iv Chapter I. II. III. IV. TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . Achievement Motivation Research and women 0 Q C C O O C C O O C The Achievement Motivation Model. . . Fear of Failure . . . . . . Statement of the Problem . . . . . METHODOLOGY . . . . . . . . . . Subjects . . . . . . . . . . Materials . . . . . . . . . . Group Assignment . . . . . . . . Session II Materials. . . . . . . Procedure . . . . . . . . . . First Session . . . . . . . . . Second Session. . . . . . . . . Numerical Data and Data Analysis. . . RESULTS. . . . . . . . . . . . Sex Role Orientation. . . . . . . Hostile Press Imagery . . . . . . The Experimental Hypotheses . . . . Validity Measures. . . . . . . . Scrambled Words Data. . . . . . . Level of Aspiration . . . . . . . The Relationship Between Fear of Failure and Fear of Success . . . . . . Informational Data . . . . . . . Summary . . . . . . . . . . . DISCUSSION. . . . . . . . . . . Task Congruence . . . . . . . . Fear of Failure . . . . . Preference for Task Difficulty . . . Level of Aspiration . . . . . . . Level of Expectation. . . . . . . Page Fear of Failure vs. Fear of Success . . . 107 Fear of Failure in Women . . . . . . 108 Fear of Failure: A Theoretical Con— ception . . . . . . . . . . . lll Implications and Directions for Future Research . . . . . . . . . . . 114 Summary . . . . . . . . . . . . 115 APPENDICES Appendix A. CUE INTERPRETATIONS . . . . . . . . . 118 B. PERSONAL CHARACTERISTICS QUESTIONNAIRE. . . 123 C. SCRAMBLED WORDS TEST. . . . . . . . . 128 D. DATA QUESTIONNAIRE . . . . . . . . . 136 E. SESSION I INSTRUCTIONS . . . . . . . . 138 F. SCRAMBLED WORD INSTRUCTIONS . . . . . . 140 G. POST-PERFORMANCE ESTIMATES. . . . . . . 141 H. PREDICTION OF ABILITY ON TESTS OF CAREER AND HOMEMAKING ABILITY . . . . . . . 150 I. ATTRIBUTION OF PERFORMANCE. . . . . . . 162 J. HISTORY OF THE PROBLEM . . . . . . . . 176 EFERENCES . . . . . . . . . . . . . . 194 vi LIST OF TABLES Answer Distributions--Questions Used for Sex Role Assignment . . . . . . . . Answer Distributions--Questions Not Used for Sex Role Assignment . . . . . . . . Three-Way Anova--Level of Aspiration . . . Three-Way Anova--Aspiration Level Minus Expectation Level. . . . . . . . . Four-Way Anova-—Anagram Task Performance. . Three-Way Anova--Anagram Task Performance . Anagram Total--Means and Standard Deviations Three-Way Anova—-Interval l Anagram Scores . Three-Way Anova--Interval Two Anagram Scores . . . . . . . . . . . . Three—Way Anova--Interva1 Three Anagram Scores. Three-Way Anova—-Interva1 Four Anagram Scores Three-Way Anova--Interval Five Anagram Scores Three—Way Anova-—Interval Six Anagram Scores Two—Way Anova--Anagram Score——Career Test . Two-Way Anova—-Anagram Score-—Homemaking Test Two-Way Anova—-Genera1ized Preference for Task Difficulty . . . . . . . . . Level of Aspiration--Means and Standard Deviations . . . . . . . . . . - o Page 32 41 48 50 54 55 56 57 58 59 60 61 62 64 65 67 69 Table 18. 19. 20. 21. 22. 28. 29. 31. Three-Way Anova-—Level of Expectation . . . Level of Expectation-~Means and Standard Deviations . . . . . . . . . . . Hostile Press Imagery Scores, Fear of Failure Scores, and Correlations . . . . Correlations Between Hostile Press Imagery Scores and Components of the Fear of Success Scoring System . . . . . . . Correlation Matrix--Variab1es Subject to Analysis of Variance . . . . . . . . Three-Way Anova-—Numerical Anagram Estimate . Three-Way Anova——Percentile Anagram Estimate . . . . . . . . . . . . Three-Way Anova--Difference Between Anagram Performance and the Numerical Post- Performance Estimate . . . . . . . . Three-Way Anova--Estimated Percentile——Career Ability Test. . . . . . . . . . . Estimated Career Ability--Means and Standard Deviations . . . . . . . . . . . Three-Way Anova--Estimated Percentile—— Homemaking Ability Test . . . . . . . Estimated Homemaking Ability-—Means and Standard Deviations . . . . . . . . Three-Way Anova--Difference Score: Estimated Percentile on Test of Homemaking Ability and on Test of Career Ability . . . . . Three-Way Anova-—Absolute Value of Dif- ference Score: Estimated Percentile on Test of Homemaking Ability and on Test of Career Ability . . . . . . . . . Three—Way Anova--Attribution to Effort. . . viii Page 70 71 72 74 78 142 143 145 151 152 154 155 156 157 166 33. 34. 35. 36. 37. 38. 39. Attribution to Effort-~Means and Standard 0 o o a 0 Deviations . . . . . Three—Way Anova-—Attribution to Skill. Attribution to Skill-—Means and Standard Deviations . . . . . . . . Three-Way Anova--Attribution to Luck . Attribution to Luck--Means and Standard Deviations . . . . . Three-Way Anova--Attribution to.Task Dif- ficulty . . . . . . . Attribution to Task Difficulty--Means and Standard Deviations . . . . . . ix 167 169 170 171 172 173 175 CHAPTER I INTRODUCTION Achievement Motivation Research and’Women The exploration of achievement motivation has gaged many personality researchers over the last 20 to 5 years. Most of this research has employed male sub— :cts, but within the last 5 years there has been a :ritable explosion of articles concerning the achievement Itivation of women. Although experimenters initially tempted to verify the McClelland—Atkinson model of hievement motivation, most recent research has attempted explain the failure of White women in American society realize their intellectual potential in conventionally :ognized ways. The manifestations of this failure are obvious ein & Bailey, 1973). In comparison to men of equal lity, fewer women enter and complete college, earn anced degrees, or use their degrees once they are ned. At each level of academic achievement the pro- tion of women decreases (Bardwick, 1971). Even within field of psychology, which has a relatively high roportion of female members (24%), women publish fewer rticles than men (Teghtsoonian, 1974). This finding is onsistent with Terman's (1959) discovery that gifted omen who enter academic fields publish significantly ewer articles and receive significantly fewer promotions an men at the same ability level. One popular explanation for this situation is that st women perceive success in intellectual and leadership Etivities as unfeminine and, therefore, have a fear of iccess (Horner, 1968; 1971a; 1972a; 1972b; 1973; Horner Walsh, 1973). The concept of this motive to avoid wcess was designed to fit into the McClelland—Atkinson del of achievement motivation. While the validity of is model for women has recently been questioned (Stein Bailey, 1973), it is the contention of this author that adaptation of this model, as suggested by Veroff (1969) i Feather (1965a), provides a more adequate model for a achievement behavior of both men and women. An :ensive discussion of the achievement motivation liter- Ire concerning women is contained in Appendix J. Why study achievement motivation? Numerous dies using male subjects have demonstrated relationships ween Achievement imagery in stories told to T.A.T.-type uli and various measures of performance. McClelland, 'nson, Clark, and Lowell (1953) report significant tionships for males between n Achievement and number f words written in an essay, number of Rorschach responses iven in a testing situation, number of words unscrambled consecutive intervals, college grades, and S.A.T. ores. It was found that the behavior of high and low Ach subjects differed when they performed tasks under utral, task—oriented, and achievement-arousing con- tions. Furthermore, achievement-arousing instructions re found to increase the amount of achievement imagery [educed by subjects. I Several recent reviews (Klinger, 1968; Klinger & Nelly, 1969; Entwisle, 1972) challenge the validity of e imagery-based measure of n Achievement (and thus the ole achievement motivation literature) because of its 3r reliability and failure to correlate with question- .re measures of achievement motivation. Inconsistencies the literature are also cited as evidence of the :stionable validity of the positive findings. Essential the consideration of these inconsistencies is the ermination of whether the T.A.T. n Achievement measure Esses the strength of the achievement motive (which Id be consistent across conditions for an individual) he strength of resultant achievement motivation ch should vary across conditions). Likewise, this ure may assess intensity of motive (or motivation) xtensivity (the degree of generalizability). It is the position of this author that the T.A.T. easure is sensitive to resultant motivational strength ather than motive strength. If this were not the case, dividual S's would not demonstrate increases in n Achieve— nt imagery from neutral to "achievement-aroused" con- 'tions. Since such changes are consistently found, it ems reasonable to conclude that the poor reliabilities e a function of the values of the other components of e resultant achievement motivation equation. But, in liS explanation we have placed the cart before the horse. t us first examine the theory of achievement motivation. The Achievement Motivation Model Although many others have contributed to formu— tion of the achievement motivation model, basic credit at be given to John Atkinson and David McClelland. In air 1953 book, The Achievement Motive, the motive was rst extensively defined and described. Motive was here ’ined as "the reintegration by a one of a change in an ective situation (1953, p. 28)." By 1958 Atkinson had fted the emphasis of this definition away from a change affect. Motive was now defined "as a disposition to ive for a certain kind of satisfaction, as a capacity satisfaction in the attainment of a certain class of antives (1966, pp. 12-13)." The achievement motive defined as the disposition to strive for satisfaction ttive to a standard of excellence. From the beginning cClelland and Atkinson recognized that resultant achieve- ent motivation is a function of both an approach and an voidance component. De Charms and Dave (1965) call iese components hope of success and fear of failure. In 1eir earliest discussion of the motive to approach success d the motive to avoid failure, McClelland, Atkinson, et . (1953) assumed that these two components of resultant thievement motivation have an additive effect on resul- nt motivation. Later, after more data were collected, kinson (1958) hypothesized that motivation to avoid ilure has an inhibitory effect on motivation to approach :cess and thus reduces the intensity of resultant nievement motivation. The McClelland-Atkinson model is called an >ectancy—value theory of motivation. Within this :ory, motivation is considered to be a function of the .tiplicative combination of motive strength, expectancy motive attainment, and the incentive value of attain- t. Atkinson (1966) hypothesizes that the incentive ue (of success or failure) equals one minus expectancy a subjective probability of success or failure). Fur— rmore, subjective probability of failure is believed 3e equal to the negative value of the subjective >ability of success. As a consequence of these Ithesized mathematical relationships, resultant .evement motivation is believed to be maximal when 6 jective probability of success equals .5. In uality, high need achievement individuals seem to fer objective probabilities of success which are some- t lower than those predicted by the equation (Atkinson & ther, 1966; Horner, 1968). The validity of these ntitative relationships between subjective probability success and incentive value has been questioned by ther (1965a), Veroff (1969), and Birney, Burdick, and an (1969). Resultant achievement motivation is also believed 5e a function of the strength of motivation to avoid :ess (Horner, 1968) and of extrinsic factors such as ffiliation or financial incentives in addition to .vation to approach success and motivation to avoid .ure (Atkinson & O'Connor, I966). The motive to avoid ess is conceived by Horner as a tendency to avoid ess because of its aversive qualities; it is defined ombining multiplicatively with expectancy and incentive = to yield motivation to avoid success. An achievement .a Iation equation can, therefore, be written TA — - AF) — M_s + ext when TA = the tendency to achieve, motivation to succeed, MAF = motivation to avoid Ire, M_S = motivation to avoid success, and ext = nsic factors. Heinz Heckhausen, a German author, has developed cry of achievement motivation which is similar, but : identical, to the McClelland-Atkinson model. He .15 the expectancy-incentive component "the expectation Idient (1967, p. 3)" and places a relatively greater Ihasis on time and other factors as determining both tectancy and incentive. In addition, he suggests that :standard of excellence in achievement motivation may either internal or external (competitive). This idea is emphasized even more by Joseph Veroff 69). Achievement motivation in the adult is conceived by him, as an integration of competency motivation social comparison. Within this conception the incen- e value of success, and, ultimately, resultant moti- ion to approach success, can be high only if the )abilities of achieving success on the task and of :ifying social needs within the context of the task F a high positive correlation. It seems to this author that Veroff‘s conception inates the necessity of Horner's (1968) motive to d success. If, as Horner suggests, success in llectual and leadership activities has an aversive 1 effect for women, then the incentive value of $3 at these activities is reduced, with a conse- reduction in motivation to succeed. If failure at lectual and leadership activities has less aversive l consequences for women than it does for men, then ncentive value of failure at these activities is er for women than it is for men. If motivation to avoid success is eliminated from a resultant achievement motivation equation, we are :ed, once again, with hope of success, fear of failure, 1 extrinsic factors. The focus of this study is to Imine both hope of success and fear of failure on women. 7 primary focus is on fear of failure, which has :eived relatively little attention in the literature cerning achievement motivation in women. Fear of Failure Motivation to avoid failure was, at first (McClel- d, Atkinson, Clark, & Lowell, 1953), believed to facili- e performance but is now (1966) considered by Atkinson al. to have an inhibitory effect on performance. Birney, dick, and Teevan (1969) retain the additive model with addition of the Yerkes-Dodson hypothesis. That is, 3 suggest that resultant achievement motivation teases as an additive function of motivation to oach success and motivation to avoid failure up to maximal point and thereafter begins to decrease. All three models (facilitory, inhibitory, and es—Dodson) assume some connection between fear of are and anxiety. Atkinson (1966) suggests that Jation to avoid failure may be associated with aty but that the anxiety may not be experienced unless .s greater than M Heckhausen (1967) posits that S- :ty will be experienced in association with the preach—avoidance conflict produced when MAF = MS' The rkes—Dodson hypothesis, which is favored by Birney, rdick, and Teevan (1969), would suggest that anxiety uld appear when both MAF and Ms are high. In other rds, when the additive combination of MS and MAF exceeds e point of maximal constructive arousal, this arousal experienced as anxiety which inhibits performance. The question of when anxiety becomes salient is cial if one attempts to measure fear of failure. The st popular mode of assessment is administration of :iety questionnaires, either the Test Anxiety Question- .re of Mandler and Sarason (1952) or the Achievement iety Questionnaire of Haber and Alpert (1960). For Atkinson formulation the questionnaires should be quate in differentiating between individuals who are e motivated by fear of failure than by hope of success. the other hand, the anxiety questionnaire technique 11d prove inadequate in assessing the absolute strength fear of failure since it is only the surplus of fear of .ure over hope of success which is experienced as ety. The questionnaire technique is wholly inadequate both the Heckhausen and Birney, Burdick, and Teevan ulation. I If the questionnaire technique of assessment of of failure is abandoned, analysis of projective es remains as another option. Numerous attempts have 10 I made to provide a story analysis technique of ssing fear of failure. De Charms and Dave (1965) Heckhausen (1967) scored achievement stories as indi- ng either hope of success or fear of failure, depending heir content and outcome. The hostile press scoring 3m developed by Birney, Burdick, and Teevan (1969) as not only achievement-related fears but also "the .le threat of non-achievement forces (1969, p. 156)." Ichievement-related scoring systems classify only .cles and task failure as indicating fear of failure, the hostile press system includes several categories terpersonal sanctions. For our formulation, which es that incentive value is a joint function of task rns (relevance to personal goals, responsibility for :mance, and probability of success) and interpersonal :ns, the hostile press system is far more adequate. Let us reiterate the author's conception of 'ement motivation. We essentially retain the land-Atkinson expectancy-value theory but without pothesized mathematical relationship between sub— e probability of success and incentive value of s. The incentive value of success, we believe, has at two determinants. The incentive value of success trticular task is determined, in part, by the rele— >f success on the task to the individual's personal :ion of competency. At the same time, incentive 11 e is affected by the implications of task success for ification of the individual's interpersonal needs. e are many possible variants of personal definition ompetency. For example, a woman may value good grades chool and completion of a training course without ing promotion to a leadership position in her field. e same token, she might value being a good cook and petent mother without caring about the appearance of Fr herself or her house. For one individual, success on the same task may different incentive values depending upon the social .cations of success. For example, suppose a pro— onal woman writes a book. Since she has defined 1f as a professional woman, publication of the book d have a high incentive value for her sense of nal competence. If her family and peer reference value professional achievement for women, the inter- 1a1 incentive value will also be high. Suppose the ' and peer group perceive the same success as "opriate to their expectations for a woman. Then terpersonal incentive value of success will be low, n negative. Hope of success will be greater for rst woman than for the second. Let us examine the fear of failure component for he women. For both women, failure to complete the ill have an adverse effect on their personal sense 12 ompetency. Interpersonally, however, it will have arent effects. The woman whose peer group supports assional achievement for women will lose social m if she does not complete the book, while the whose peer group does not believe such achievements ppropriate for women may increase in social esteem e stops writing. The incentive values of success f avoiding failure are clearly high for the first and lower for the second. If we add into our formulation the concept of Lve weightings of the competency and social compari— >mponents, the picture becomes even more complicated. 1e first woman only the strength of the total motive 's because both she and her reference group value 3 and abhor failure equally. The second woman's tion is more complicated. If she values personal ency more than she values social esteem, she may t in completing the book. If, on the other hand, ues social esteem over personal competency, her tendency to complete the book will be low. , if she values social esteem and personal com- equally, her approach and avoidance tendencies equal, producing a conflict of the type which sen (1967) suggests will produce anxiety. Let us shift our frame of reference for a moment. ern which we have just described for the first typical for males in our society. That is, 13 ssional success by a man is likely to be associated increased social esteem. The second pattern, in .professional success is associated with small ‘ases in social esteem or even loss of social esteem, re typical for women. It is clear that, as a group, hould be more motivated to approach professional 55 than women. Up to now we have ignored subjective probability ccess as a component in resultant achievement moti— n. Several studies (Crandall, 1969; Feather, 1969; & Feather, 1973; Feather & Simon, 1974) indicate women underestimate their ability to succeed at a ty of tasks, both in experimental and in natural :ions, while men consistently overestimate their :ies. This difference, alone, suggests that :ant achievement motivation for men should be r than resultant achievement motivation for women e women's expectations of success are lower than of men. The advantage of this conception of achievement :ion is that it explains the achievement behavior . sexes on intellectual and leadership tasks using cepts of motivation to approach success, motivation 3 failure, and extrinsic factors. This approach the hypothesized mathematical relatiOnship probability of success and incentive value of 14 cess as well as the necessity of including the motive avoid success as a separate component in the resultant Lievement motivation equation. We do not deny that some ten seem to fear success in traditionally male activi— IS. The apparently fear-motivated behavior of such ten, however, is simply behavior which is consistent .h their personal definitions of competency. If a woman :ls that success in traditionally male activities is onsistent with her personal definition of femininity, does not avoid such activities because she is afraid them but simply because she is not interested. There probably a small minority of individuals, both male female, who fear success in general because they a punished for successes in childhood. The number of Lviduals who were subjected to such pathogenic early ariences is probably low. The failure of most women .chieve success on the intellectual and leadership 5 appropriate to their intellectual potential is a tion of restrictive societal prescriptions for apriate female behaviOr, personal definitions of atency which place little emphasis on success at 1 tasks, and the tendency of most women to under- ate their abilities. Women are not, as Veroff (1969) sts, suffering from an ambivalent attitude towards 55 produced by an unsuccessful resolution of the ll triangle, but, instead, are suffering from the 15 gative aspects of the female sex role stereotype roverman, Vogel, Broverman, Clarkson, & Rosenkrantz, 72). Statement of the Problem The previously described formulation of achieve- nt motivation is far too complicated to be tested in a ngle experiment. This experiment tested one portion the formulation: That incentive value of success and, timately, performance level are dependent upon the cor- spondence between the task and the subject's personal Einition of competency. For this purpose the subjects :e divided into two groups, one of which placed rela- rely more value on success in a career and the other of .ch placed relatively more value on success in marriage , as a homemaker. Each of these groups was asked to plate a scrambled words test which was described as a t of career ability to some subjects and as a test of emaking ability to others. For purposes of brevity, jects will be described as serving in a congruent lition when their personal competency definition and task description were similar and as incongruent when personal definition of competency and the task ription were dissimilar. The theory predicts that acts will perform significantly better on a congruent than on an incongruent task. 16 Although several previous studies have examined e aspects of the problem to some extent, the relation— of performance to the correspondence between personal etency definition and task description has never been uately evaluated. Milton (1959) found that men per— significantly better than women on problems with cally male content, while women perform significantly er than men on problems with typically female content. id not, however, examine intra-sex variations. French Lesser's (1964) findings were confounded by experi- al error. Parker (1971) and Makosky (1972) related ormance to fear of success score but did not analyze r data using sex role ideology as a subject variable. all and Gonzalez (1972) did not measure performance in : study but only offered their subjects a choice of ;ble story outcomes. While all these studies lead us alieve that a subject will perform better when a task :5cribed as being congruent to her personal definition mpetency than when it is not, this hypothesis has been precisely tested. A further discussion of the 'ments described above is contained in Appendix J. In our discussion thus far, only the hope of 5 component of the resultant achievement motivation on has been considered. In this experiment we also ed the comparative effect of fear of failure moti- on performance, level of aSpiration and expectation, 17 formance estimates, and performance attribution. Ie the study of fear of failure has attracted con— erable attention in the research literature on EEAEE' ias been largely ignored by the achievement motivation ariments working with female subjects. Only nine of 60 studies of fear of failure reviewed by Birney, Bur— :, and Teevan (1969) included women; only 6 of these ies were published; and none took the possibility of differences into consideration. In his only published I y concerning sex differences in fear of failure (Smith evan, 1971), Teevan found that hostile press imagery negatively correlated with general self—ideal con- 1ce for both men and women. While there was a nega- relationship between hostile press imagery and rsement of the achievement related items for men, relationship did not hold up for women. Several studies examining the effect of fear of re motivation on the behavior of women have been shed in the last few years. A Norwegian study me, 1973) found no significant differences in the I L grades of seventh-grade girls with the four I %e combinations of n—Ach and test anxiety. Another I. . . . gian study (Vollmer, 1973) found a Significant I Tve relationship between fear of failure, assessed % Heckhausen system, and a Holtzman ink blot measure A fisonality differentiation. Crossing the globe, I A H1 Leah 18 pping over the U.S.A., we find that a Hong Kong study ng-Fun Li, 1974) discovered significant positive cor- ations between low test anxiety in fifth— and sixth- ie girls and reported parental attitudes of dominance harshness as well as of encouragement of communi— .on, comradeship, and sharing. Only two American studies examining variations in ‘of failure among women seem to have been published. er (1968) found no significant relationship between of failure (as measured by the Debilitating Test ety subscale of the Haber—Alpert Achievement Anxiety tionnaire) and performance on several measures of al and analytic skill. Karabenick and Marshall (1974) :ed test anxiety (as a measure of fear of failure), of success, competition condition, and type of feed- to improvement on a digit substitution task. A sig- ant three-way interaction was discovered between fear ccess, opponent condition, and test anxiety. Low of success subjects with low test anxiety improved when competing against a man, while those with high anxiety improved most when competing against a woman. fear of success subjects with low test anxiety red most when working alone, and those with high Inxiety improved most when competing against a man. found that low test anxiety subjects improved most 19 :r failure feedback, while high test anxiety subjects 'oved most after success. Obviously, the literature on fear of failure moti— on in women is meager and in need of extension. In study we contrasted the behavior of subjects who ed above and below the median on the hostile press ery (HP) measure of fear of failure. We did not 3y the test anxiety measure of fear of failure which )een used in the previously described studies (Daugert, ' Horner, 1968; Karabenick & Marshall, 1974) which .ned the relationship between fear of failure and rrmance. The inhibitory model of fear of failure, predicts that resultant achievement motivation will creased by the magnitude of motivation to avoid re, was the basis for our predictions. The combi- n of task congruence and fear of failure produces :onditions roughly akin to the four combinations of and fear of failure used in the male studies (high high fear; high n-Ach, low fear; low n-Ach, high and low n-Ach, low fear). In this experiment it sumed that task congruence stimulates a higher level e of success relative to that associated with task ruence. In line with the inhibitory model of fear of a motivation, the following predictions were made. ear of failure is low, subjects will perform better 20 :r conditions producing high hope of success than under .itions producing low hope of success. Re-phrased in s of this experiment the hypothesis becomes: Hypothesis 1: Low fear of failure subjects will unscramble sig- nificantly more words in the congruent condition than in the incongruent condition. When fear of failure is high, it should inhibit affect of hope of success on resultant achievement ration. Furthermore, fear of failure is greatest when of success is maximal. Since, in this experiment, . hypothesized that hope of success is greater in the uent condition than in the incongruent condition, as ollary, fear of failure should be greater in the con— t condition than in the incongruent condition. As sequence, the resultant achievement motivation of fear of failure individuals should be inhibited by 3f failure more in the congruent condition than in icongruent condition. In terms of this experiment, 'pothesis was stated: 'pothesis 2: gh fear of failure subjects will unscramble sig— ficantly more words in the incongruent condition an in the congruent condition. 21 Finally, the model predicts that when hope of cess is high, subjects with low fear of failure should form better than subjects with high fear of failure. this experiment, the hypothesis became: Hypothesis 3: In the congruent condition, low fear of failure subjects will perform significantly better than high fear of failure subjects. If these hypotheses are confirmed, we can tenta— ly conclude that fear of failure has an inhibitory ct on resultant achievement motivation in women when ess on the task is likely to be instrumental to evement of a goal valued by the subject. This con- Lon is consistent with the findings of studies which ayed male subjects. High fear of failure males con- :ntly (Feather, 1965a; Hancock & Teevan, 1964; Thomas :van, 1964) prefer tasks which are either extremely 1or difficult, while low fear of failure males choose I at intermediate levels of difficulty. The women in experiment were asked to estimate their generalized Fence for difficulty of task. Hypotheses 4 and 5 Lted replication of the male study results. I Ipothesis 4: I ‘gnificantly more subjects with low fear of failure Ian subjects with high fear of failure will prefer Itermediate task difficulty (30 to 70% probability I success). I I I 22 Hypothesis 5: Significantly more subjects with high fear of failure than subjects with low fear of failure will prefer extreme task difficulty (0 to 29 and 71 to 100% probability of success). We also asked our subjects to indicate (in per- .tiles) their level of aspiration on the scrambled words k. High fear of failure individuals typically (Birney, dick, & Teevan, 1969) set their level of aspiration remely high or extremely low relative to the level of iration preferred by low fear of failure individuals. inhibitory model of fear of failure predicts that for Lviduals with fear of failure when hope of success is .mal, fear of failure should also reach its peak. It predicted, therefore, that high fear of failure sub- .s would display the defensive behaviors of aiming low ; iming high when hope of success is high. Subjects low fear of failure should aspire for an intermediate I of performance because this level of aspiration hizes the probability of confirming their expectations. {hope of success is relatively high (in the congruent tion), subjects with low fear of failure should e to intermediate levels of performance to a sig— antly greater extent than subjects with high fear of re. Phrased in terms of this experiment, the esis became: Hypothesis 6: In the congruent condition, significantly more sub- jects with low fear of failure than subjects with high fear of failure will aspire to intermediate levels of performance (30th to 70th percentile). When hope of success is high, subjects with high evels of fear of failure should aspire to extreme levels E performance significantly more often than they aspire 3 intermediate levels of performance. We hypothesized: Hypothesis 7: In the congruent condition, high fear of failure subjects with a homemaking orientation will aspire to extreme levels of performance (0—29 percentile and 71-100th percentile) significantly more often than they aspire to intermediate (30-70th percen- tile) levels of performance. A more directional hypothesis was made for the ;h fear of failure career oriented women. High fear of .1ure males have nearly the same preference for the ategies of setting level of aspiration extremely high extremely low (Birney, Burdick, & Teevan, 1969). We athesized that a high fear of failure woman who defines personal competency in terms of success at a career .ikely to set her level of aspiration extremely high. gh fear of failure woman is in a sense "aiming high" he elects to pursue a career. Discriminatory ssion, hiring, and promotion policies, along with 11 pressures, make it difficult for a woman to pursue reer. A high fear of failure, career-oriented woman 24 who prefers the strategy of aiming low is, therefore, .ikely to drop her efforts to pursue a career. Among Iareer-oriented women, then, high fear of failure indi— iduals should aspire significantly higher than low fear f failure individuals because of their preference for the aiming high" strategy. Phrased in terms of this experi— ent, the hypothesis becomes: Hypothesis 8: In the congruent condition, career-oriented women with high fear of failure will aspire significantly higher than career-oriented women with low fear of failure. Several studies (Birney, Burdick, & Teevan, 1969) dicate that high fear of failure males have less confi— ace in their performance than low fear of failure sub— :ts. Lack of confidence in the high fear of failure I is likely to be expressed in large differences between e1 of aspiration and level of expectation. We there— e predicted: Hypothesis 9: The difference between performance aspiration and performance expectation will be significantly greater for the high fear of failure subjects than for the low fear of failure subjects. In addition to those variables for which hypothe— iave been presented, a number of other relationships examined. Subjects were asked to estimate the 25 umber of words which they had unscrambled, and this stimate was compared both to their performance and to he estimates which they made prior to attempting the ask. In addition, they were asked to estimate their erformance level in percentiles. They were also asked 0 predict their performance level on a test measuring ility to be a homemaker and on a test measuring ability succeed in a career. Our interest here was to discover ether subjects differing in level of fear of failure tivation also differ in the extent to which they assume at the ability to succeed on one type of task precludes Le ability to succeed on a task of the opposite sex role ientation. Finally, the relationship between Horner's w scoring system for the motive to avoid success and e hostile press scoring system for fear of failure was amined. CHAPTER II METHODOLOGY Subjects The subjects were 100 female introductory psy— 1ology students at Michigan State University drawn from initial (first session) sample of 156 subjects. The I periment was conducted in October and November of 1973. ghty-three individuals volunteered for the experiment .t did not attend the first session. Thirty-three sub- cts were excluded from the experiment because they did t have a clear preference for one sex role orientation. 2 two subjects who were born in foreign countries were :luded. Using the criteria listed in the Group Assignment tion, 69 subjects were regarded as homemaking oriented; were regarded as career oriented; and 33 were con- sred to be undecided about their sex role orientation. :he subjects asked to return for the second session, :areer-oriented and 58 homemaking—oriented subjects nded. As described in the Group Assignment Section, dian cut of the homemaking and career—oriented sub- 5, taken together, gave a classification of 58 subjects 26 27 fear of failure SS and 63 SS without fear of failure. these subjects, 48 subjects with fear of failure and without fear of failure attended the second session. number of subjects attending the second session who 1 into each of the four categories created by combi— ion of these two variables was: (1) Career orientation, h fear of failure (17); (2) Career orientation, low r of failure (25); (3) Homemaking orientation, high r of failure (31); and (4) Homemaking orientation, low r of failure (27). Materials During the first session, the SS were adminis— ed four story leads in the booklet entitled "Cue erpretations." (See Appendix A.) The instructions this booklet were adapted from Horner (1968) and fered frOm her instructions only in minor details. .de the booklet in random order were four story leads .ted for this experiment--one career-oriented activity, homemaking-oriented activity, and two feeling state riptions. gs were given five minutes to write each Y, with a verbal warning one minute before the time 1p. The career—oriented and homemaking—oriented Iity leads employed different settings; the homemaking placed the woman in a kitchen, while the career lead :he woman at work. 28 The career-oriented lead was: A woman is working a an office. The homemaking-oriented lead was: A woman is >rking in her kitchen. I The two feeling leads were: (1) A woman has a tile on her face; (2) A woman has a thoughtful look on Ir face. The stories composed by the SS were content alyzed for fear of failure using the hostile press agery scoring system developed by Birney, Burdick, i Teevan (1969) and for fear of success using the new 973) Horner scoring system. Horner's new scoring sys- n for fear of success requires the inclusion of both :ivity and nonactivity stimuli. She now believes that :ponses to her "Anne-medical school“ stimulus were 'hly influenced by the stimulus content. The results a pilot study conducted by this author suggest that any cific information describing the character in the nulus colors the stories told by the subjects. For 3 reason the stimuli were made as uniform as possible 1 only the activity or feeling state of the woman :ified. The hostile press scoring system for fear of ure used in this study has no significant correlation scores on either the Haber—Alpert Dehabilitating avement Anxiety Test or the Mandler-Sarason Test 29 Anxiety Questionnaire (Birney, Burdick, & Teevan, 1969) which have traditionally been used as measures of fear of failure motivation. On the other hand it has a signifi— cant negative relationship to n-Ach; this relationship is consistent with the assumed inhibiting effect of fear of failure on resultant achievement motivation. Hostile press imagery is significantly greater following failure feedback than under neutral conditions. The consistent iifferences in level of aspiration between high and low fear of failure individuals found in studies using the T.A.Q. and Haber—Alpert measures also appear when the lostile press imagery scoring system is used. A variety If other studies indicate that high scores are predictive f a pattern of behavior consistent with the conception of motive to avoid failure. Group Assignment The Cue Interpretations Booklet, which contained 1e story leads, provided for identification of gs only ' a code number. The four stories which §s wrote to the Ies in this booklet were separated and sorted into a cup with all the other stories for the same lead. ese stories were scored for hostile press imagery and r the motive to avoid success by the author. Twenty atocols were selected randomly and scored by a graduate 1dent in psychology. Interjudge reliability for :tile press imagery was 100%. An additional 20 30 rotocols were selected randomly and rescored by the uthor six months after the initial scoring. Intra- udge reliability for hostile press imagery was 95%. The number of hostile press imagery (fear of ilure) stories told by each S was computed, and a fre- ency distribution for the number of hostile press agery stories told by the subjects classified as having ther a career or homemaking orientation was completed. e median for this distribution was one; subjects telling Ie or more hostile press imagery stories were categorized having high fear of failure, while subjects telling hostile press imagery stories were categorized as ving low fear of failure. The Personal Characteristics Questionnaire (see pendix B), administered to the gs at the first session, developed for this experiment; it included questions d for determining the sex role orientation as well as uestion indicating task difficulty preference. These stions were buried among filler items, with the intent distracting the S's attention from the questions con- ning sex role orientation. Some of the filler stions yielded interesting results, but they were 'gned to convince the SS that E was interested in e variables which differentiate women who go to ege from those who do not. 31 Six questions were included to determine sex role ientation. Three of these questions were adapted from rker (1971). The other three questions were taken from mquist and Angrist (1970). Only three of these estions were actually useful for assigning subjects to e sex role orientation classification. On the basis of eir responses to these three questions, gs were desig- ted as career oriented, undecided, or homemaking iented. Assignment to either the career or the homemaking x role orientation required a relatively consistent ttern of responses to the three questions. The estions used appear in Table l. A subject was clas- Eied as having a homemaking orientation if she gave one the following patterns of response: c, d, or e to astion l3, 4 or 5 to question 15, and a, b, or e to stion 35. A subject was also classified as having a emaking orientation if she answered d or e to question 2 to question 15, and a, b, or e to question 35; she also classified as having a homemaking orientation if answered d or e to question 13, 4 or 5 to question 15, c to question 35. A subject was classified as having reer orientation if she answered a or b to question 1 to question 15, and d to question 35. A subject also considered career oriented if she answered c to tion 13, 1 to question 15, and d to question 35, or he answered a or b to question 13, 2 to question 15, 32 Table 1 Answer Distributions-—Questions Used for Sex Role Assignment If you could have only a career or only marriage, which do you think you would choose? Definitely career without marriage (12) b. Would probably prefer career rather than marriage (19) c. Undecided (45) _ Would probably prefer marriage without career (65) do e. Definitely marriage without career (16) Would you want to work under the following con- ditions? a. One child of school age, husband's salary adequate Definitely yes (54) Probably yes (60) Undecided (18) Probably not (15) Definitely not (10) mhwww Assume that you are trained for the occupation of your choice, that you will marry and have children, and that your husband will earn enough so that you will never have to work unless you want to. Under these conditions, which of the following would you prefer? To participate in clubs or volunteer work (5) To spent time on hobbies, sports, or other activities (5) To work part-time in your chosen occupation (78) To work full-time in your chosen occupation (29) To concentrate on home and family (12) a. b. C. d e. nd d to question 35, or if she answered a or b to question 3, 1 to question 15, and d to question 35. Those subjects t classified as having either a career or a homemaking ientation were considered undecided and were eliminated om further consideration. Subjects who grew up in a untry other than the U.S.A. were also eliminated. Thus, four groups of uneven size were formed. e high drop—out rate before the first session suggested at there might be a high drop-out rate between the first d second sessions, so the four groups were left uneven size. On a random basis, half the gs in each group re assigned to a congruent condition and half to an congruent condition. Session II Materials During the second session gs were administered 3 Scrambled Words Test (see Appendix C), six pages with 4—letter scrambled words on each page which were taken Im Parker (1971). On the page following the instructions the scrambled words task, gs were asked to indicate a Scale from 1 to 100, divided into five—unit intervals, hich percentile they would try to have their perfor— e to fall and in which percentile they expected their ormance to fall. After completion of the Scrambled Words Test, the ere asked to fill out the Data Questionnaire. (See ndix D.) In this questionnaire the SS were asked to mate the number of words they had unscrambled and also 34 he percentile into which they thought their performance ad fallen. Subjects were asked to indicate in per- entages how much of their performance they attributed to ffort, skill, luck, and task difficulty. Subjects were so asked to predict their performance on a test measur— g ability to be a homemaker and on a test measuring ility to be successful in a career. One of these estions provided us with the S's estimate of her ability succeed at the sex role activity which is closest to r personal definition of competence and the other estion enabled us to determine the relationship of is estimate to her estimated ability for the opposite x role orientation. Analysis of this data allowed amination of the hypothesis that high fear of failure assume that ability in one area implies ineptness in >ther. Three questions were included to evaluate the I acceptance of the experimental manipulations. On the st page of the Data Questionnaire, gs were asked, "How d a test of career/homemaking ability do you think this t was?" The choices for answers were "excellent," d," "fair," and "poor." On the second page of the tionnaire gs were asked, "What do you think this riment was about?“ and "Did you believe the experi- er when she told you that the scrambled words test elated with success in a career/as a homemaker?" 35 Procedure There were two sessions conducted in a medium— :ized classroom with about 20 SS present at a time. Two ’emale undergraduate psychology students served as experi— enters. Only one experimenter was present as each group as tested, but each experimenter ran groups of subjects n both Session 1 and Session 2. All but nine subjects ad the same experimenter for both sessions. The experi— enters, like the subjects, were led to believe that the (periment was designed to test the differences between amen who do attend college and those who do not. After Impletion of the experiment, both experimenters were ked what they thought was the purpose of the experiment. ither experimenter believed the author, and one thought e experiment was concerned with achievement motivation, vever, both were unaware of the different groups of )jects and the hypotheses concerning their behavior. gs signed up for a particular initial session of 2 experiment on sign-up sheets in their introductory 'chology classroom and were reminded by telephone a day two prior to the experiment. Prior to the second sion, gs were contacted by the author and asked to E to a second session. Subjects received research Hits for their introductory psychology course in brn for their participation. 36 First Session At the first session the E handed out the two booklets and recited the instructions (see Appendix E). gs were given five minutes to complete each story in the Cue Interpretations Booklet. This is the same period of time used by Parker (1971) and in the pilot study and is far easier to time accurately than the four minutes used y Horner (1968). There was no time limit for the ersonal Characteristics Questionnaire. Second Session After four groups of subjects were selected on :he basis of fear of failure and sex role orientation Icores as previously described, each S was contacted by he author and asked to return for the second session. efore each session E received from the author a list of he subjects who were to attend, along with the instruc— Lons she was to administer. Fifteen to 20 Se attended :ch of the second sessions. While these testing groups >ntained women of both sex role orientations, all the : in each testing group received the same condition .structions. The general instructions for the Scrambled rds task (see Appendix F) were given prior to adminis- ation of the condition instructions. §s were adminis— ed either the homemaking or career condition instruc— ns. 37 Subjects who received the homemaking condition nstructions were told, "This task is a measure of verbal acility, which, studies indicate, is related to socia— ility and empathy, two attributes which are crucial to homemaker's success." Subjects who received the career >ndition instructions were told, "This task is a measure ’ verbal ability, which, studies, indicate, is related the analytic and integrative skills necessary for ccess in a career." After E administered the achievement-arousing structions, she asked the gs to read the instructions : the Scrambled Words Test. When the SS finished 1ding, she said, "Before we begin, pleaSe answer the I questions on the second page." When the SS finished .5 task, E said, "You will have two minutes for each e. Remember, do not turn the page until I tell you go ahead. Ready, begin." I After completion of this task, E stated, "Now e is another questionnaire I would like you to com— e, but first I want to collect the materials you've ady finished." E then collected the Scrambled Words 8. Then she said, "Now I'd like you to complete the Questionnaire. There is no time limit. Please be to write your name on the front page; we need this mation so that we can relate your data from the session to the information we receive from you 38 oday. When you are done, please bring the booklet up ere to me. Thank you, again, for your participation." he Scrambled Words Test booklet and the Data Question— aire were identified by a code number so that results 1 one could be related to results on the other. Numerical Data and Data Analysis The experiment yielded considerable data. To gin with, there is the S's performance data on the rambled Words Test. Secondly, gs indicated their level preferred task difficulty as well as their level of piration and expectation just prior to taking the rambled Words Test. SS were also asked to estimate air performance numerically and in percentiles follow— I completion of the experimental task. The SS were juired to indicate their estimated performance on a task the sex role orientation opposite to their own. gs e also asked to indicate the percentage of their per- mance attributable to effort, skill, luck, and task Ficulty. An additional body of data provided general triptive information about the §S as well as infor— Ion about their acceptance of the experimental Ipulations. j Each S's Scrambled Words Test was scored and ‘ I Ihed. A 2 X 2 X 2 factorial unweighted means analysis briance was computed for this data. A similar ysis of variance was performed on the level of I I I 39 spiration and level of expectation data as well as on he level of aspiration for goal opposite task data. iditionally, a difference score indicating the numerical Lfference between the actual and self—estimated per- >rmance was computed and a 2 x 2 X 2 factorial unweighted Ians analysis of variance was computed for this data. The S's protocols were scored for both hostile ess imagery and the motive to avoid success. A coef- cient of correlation was computed between the two asures. A coefficient of correlation between each the six components of the motive to avoid success )ring system and the total measure of hostile press agery was also computed. CHAPTER III RESULTS Sex Role Orientation Seven of the questions in the Personal Charac— ristics Questionnaire were to have been used to deter- ie sex role orientation. Four of these questions did : yield sufficient variation in response so that they [Id be used to assign subjects to a sex role orientation Iup. These questions and the distributions of their wers are listed in Table 2. The skewed distributions answers to questions 7, 18, and 30 suggest that these ms may have a large component of social desirability. the other hand, this pattern can be interpreted as Tcating that most of our subjects are attracted to Imaking and a career simultaneously. If this is the I, our assignment of gs to one or the other sex role Intations is clearly a relative rather than an absolute iment. These three questions were not used to determine hole orientation because there was insufficient pbility of response. Question 15b was actually a Alary to one of the questions which was used (15a) and pnses to this question differed so little from 40 41 Table 2 Answer Distributions--Questions Not Used for Sex Role Assignment How important do you feel that a professional career of your own is? Important (126) Probably important (20) Undecided (9) d. Probably unimportant (2) Unimportant (0) Would you want to work under the following conditions? Two or more children of school age, husband's salary adequate. QU‘QJ (‘D l. Definitely yes (51) 2. Probably yes (63) 3. Undecided (l6) 4. Probably not (17) 5. Definitely not (9) How important do you feel that marriage and a family are to you? a. Important (108) . Probably important (32) c. Undecided (17) d. Probably unimportant (5) e. Unimportant (5) U: Fifteen years from now would you like to be: a. A housewife with no children (0) b. A housewife with one or more children (25) c. An unmarried career woman (4) d. A married career woman without children (15) e. A married career woman with children (127) f. Other (0) 42 answers to question 15a that its inclusion would have had little effect on subject assignment. Question 30 appeared to tap social desirability. Questions 13, 15a, and 35 were used to assign sex role orientation. The distributions of responses to these questions are indicated in Table 1. Assignment of ubjects to sex role orientation was done on a purely ragmatic basis to provide the best three—way split of e sample. On the basis of the criteria described in e Procedure chapter, 69 subjects were categorized as omemakers, 52 were categorized as careerists, and 33 ‘ubjects were regarded as undecided about their sex role rientation. After completion of the experiment, a sex role ieology scale was developed to evaluate the parameters 3 sex role orientation for the entire sample. All six ,estions were used in this scale and were weighted ually. Answers to each question were given a value of a if they suggested a career orientation, two if they rgested a homemaking orientation, and zero if they e "undecided." The range for this scale was 3—12 h a midpoint of 9. For the sample the mean score 7.3 with a standard deviation of 1.6; the median 7.0, and the mode was 8.0. The range of scores was 11. It would appear, therefore, that the sample whole tended to be career oriented. This finding 43 is not surprising in light of the fact that the sample was drawn from a college population. An important implication of this result, however, is that our assumption of task congruence when the homemaking task was presented to the homemaking oriented gs may be invalid. Hostile Press Imagery The stories composed by the subjects were scored or hostile press imagery using the Birney, Burdick, nd Teevan (1969) system with a maximum possible score f 6 for each story. The four cues, in spite of their pparent neutrality of content, were markedly different n the amount of hostile press imagery which they stimu- ated. Mean hostile press imagery scores for each lead ad their standard deviations were as follows: Kitchen, = 1.040, c = 1.860; office, 7: = .390, o = 1.113; Ioughtful look, 2 = .710, o = 1.589; and smile, i = 80, o = .753. There was a marked positive skew to 1 these distributions. Thirty—one hostile press Igery stories were told to the "kitchen" stimulus, to the "office" stimulus, 22 to the "thoughtful look" nulus, and 8 to the "smile"-stimulus. The most common rce of the need press in the "kitchen“ stories was e, hostile environment (18 stories). The majority hese stories were reminiscent of the television arcials in which a busy housewife is simultaneously 44 esieged by a crowd of ravenous children, a telephone all, a delivery man, and a cooking emergency. In the 'office" stories, on the other hand, the most common ource of press was reprimands for personal activity, sually from a male supervisor. In both the "smile" and e "thoughtful look" stories, deprivation of an affilia- ive relationship was the most common source of need ress. Nearly half the subjects wrote at least one hos— '1e press imagery story, while slightly more than half d not write any hostile press imagery stories. Sixty— ree of the 100 subjects wrote no stories containing >stile press imagery, 39 gs wrote one hostile press lagery story, 16 gs wrote two, and 3 gs wrote three. subject wrote four stories containing hostile press agery. The mean number of hostile press imagery ories was .66, while the median was slightly below = The numeric scores for hostile press imagery pro— '- :ed a higher mean (2.30), but the median score was '11 slightly below one. Subjects who told stories with any hostile press gery were, therefore, classified as high fear of lure subjects relative to the other subjects, while se who had not told stories containing hostile press ery were regarded as low fear of failure subjects tive to the other subjects. 45 A t-test revealed a significant (p < .02) dif- ference between the mean number of hostile press imagery stories written by homemakers (X = 1.72) and the mean number written by careerists (Y = .81). Later in this chapter we shall examine the relationship between hostile press imagery scores and fear of success scores, but let 5 proceed to an examination of the experimental hypothe— es. The Experimental Hypotheses The first three hypotheses concerned performance n the scrambled words task. The first hypothesis stated, Low fear of failure subjects will unscramble signifi- antly more words in the congruent condition than in the ncongruent condition." The hypothesis was tested using two-tailed t—test for uncorrelated means and was not anfirmed (t = 1.385, df = 50, p > .05). Hypothesis 2 was, "High fear of failure subjects ;11 unscramble significantly more words in the incon- ruent condition than in the congruent condition." A mparison of these two groups produced significant = —2.046, df = 46, p. <.05) results but in the oppo- te direction from that predicted; high fear of failure jects unscrambled significantly more words in the gruent condition (i = 103.826) than in the incongruent dition (Y = 83.000). 46 The third hypothesis posited, "In the congruent ondition, low fear of failure subjects will perform ignificantly better than high fear of failure subjects." 115 hypothesis was not confirmed (t = —1.063, df = 47). Hypothesis 4 stated, "Significantly more subjects .th low fear of failure than subjects with high fear of .ilure will prefer intermediate task difficulty (30 to % probability of success). Intermediate levels of task fficulty preference were indicated by 26 out of 48 bjects with high fear of failure and 23 out of 52 bjects with low fear of failure. Although the trend .. 3 opposite to prediction, the probability that the effect ; produced by chance was greater than .5 (X2 = .18). Hypothesis 5 stated, "Significantly more subjects :h high fear of failure than subjects with low fear of Ilure will prefer extreme levels of task difficulty to 29 and 71 to 100% probability of success)." ‘ference for extreme levels of task difficulty was icated by 22 out of 48 subjects with high fear of Rure and 26 out of 52 subjects with low fear of Iure. Once again the trend was opposite to pre- tion, but the chi square value of .33 had a chance Sability of greater than .5. I Hypotheses 6 through 8 concerned the subjects' II of aspiration on the anagram task. The subjects ;been asked to indicate in which percentile they 47 would try to have their performance fall. Hypothesis 6 stated, ”In the congruent condition, significantly more low fear of failure subjects than high fear of failure subjects will aspire to an intermediate level of per— formance (30th to 70th percentile)." This hypothesis was not confirmed (X2 = .5, p < .25); in fact, only one subject aspired to an intermediate level of performance is here defined. As will be later discussed, all the >ther subjects aspired to an extremely high level of terformance. Table 3, the summary table for a three-way nalysis of variance using the unweighted means tech- ique, indicates that the task congruence by fear of ailure interaction for this data had no significant Efect. The seventh hypothesis stated, "In the congruent Indition, homemakers with high fear of failure will pire to an extreme level of performance (0-30 and -100th percentile) significantly more frequently than ey will aspire to an intermediate level of performance Lst to 69th percentile)." Once again, the hypothesis I not confirmed because all the subjects had aspired what had been defined as an extreme level of iration. Hypothesis 8 states, "In the congruent condition, Ber—oriented women with high fear of failure will .re to a significantly higher level of performance 48 oo.ooH 6mm.hma mm oav.ovm~mH HMflOB Ha.hm Mbo.mva Nm www.mmm‘ma HOHHm ow. m. Hm>o NNv. hmm.mm H www.mm .0.9 N .m.m x .m.m mm. m. Hwbo Haw. HNm.ow H Hmm.om .U.B x .h.m Ho. m. H0>O HHO. mmm.H H mmm.H .U.B X .m.m mo.w #Hvo. mam.w www.wmw H hum.wmm .m.m N .m.m mm.m Nwo. mhm.m www.mmm H www.mmm wocwflumcou me5 mm. m. Ho>o mew. mmm.ae H mmm.ns annanmm no name wm.w usmmo. mom.v mm¢.Nmm H mmv.Nmm mHOm xwm mfimwm .33 m omwmm a mwfiw Ill! cowumufimmfl mo Hmbmqllm>os¢ mmzlowuse m wanna 49 Lan career-oriented women with low fear of failure." .is hypothesis was not confirmed (t = 1.61, df = 20). The last hypothesis concerned the difference tween the percentile for which g said she would try d the percentile in which she expected her performance fall. The ninth hypothesis proposed, "The difference tween performance aspiration and performance expectation 11 be significantly greater for the high fear of ilure subjects than for the low fear of failure sub- :ts." Table 4 summarizes the results of a three-way LlYSiS of variance using the unweighted means technique ‘ the try for percentile minus the expect percentile a. Inspection of the table reveals that there was no nificant main effect for fear of failure, thus indi— ing that Hypothesis 9 was not confirmed. (Note: This all other analyses reported in this study were com- ed using the "Data Text" program on an I.B.M. 360 ’uter at the University of Chicago Computation er.) We end this section with the depressing conclusion none of the hypotheses were confirmed. There are >ossible explanations for this absence of corres— ance between hypotheses and data. On one hand, we ave examined experimental variables which have no ance to the dependent variables. 0n the other hand, {perimental variables may be relevant but our 50 \IIIIIIIIIIIIIII. 00.00H mmm.mmH mm mmm.osv.mH Hmuoe mm.mm HHm.mmH mm mmm.omm.VH uouum co. m. um>o woo. mmm. H mmm. .u.a x .m.m x .m.m Hm. m. um>o mom. mom.Hm H mom.Hm .o.a x .m.m am. new. vmm. mmv.mm H mmv.mw .o.e x .m.m m~.m who. mmH.m mmm.>¢v H mmm.smv .m.m x .m.m Hm.H emH. mm>.H mmH.ow~ H amH.om~ monwsuwaoo xmme as. mam. «ms. mav.eHH H me«.VHH musHHmm «0 Home mm. New. mmm. «vs.Hm H Nee.Hm mHom xmm Hm>wq sowuwuoomxm MUCH: Hw>oq soflpmuwmm¢11m>oc< mmzlmmnaa v manna 51 {potheses based upon an inaccurate understanding of the recesses involved. That the later explanation is more :propriate is suggested by the fact that further alysis of the data revealed numerous significant fects. we will shortly discuss these results, but st let us investigate the validity of our experimental ipulations. Validity Measures A number of measures were included in the experi- 1t to test the validity of the experimental manipu- :ions. The subjects were asked whether they believed. experimenter when she told them that the anagrams k measured career or homemaking ability. It is sur- sing that there were many significant effects found the data (which are discussed in the Discussion and =ndices) relating to the anagram task because only f the 100 subjects believed the experimenter. ty-two subjects said they did not believe the rimenter, and 46 stated that they believed her what. There was no experimenter effect on belief in iption of the task. While not significant (p < .059), was a strong tendency for the test to be perceived latively better when it was described as a career than when it was described as a test of homemaking ty. When gs were asked to evaluate the quality of est, they were similarly unimpressed. One subject 52 [ht the test was excellent, 22 thought it was good, [ought it was fair, and 31 thought it was a poor test. : was no significant association between task con— xce and/or experimenter on test quality. As would pected, there was a significant (p < .007) associ- between test quality and belief in the experimenter. In their written explanations of their belief or lief of the experimenter, many subjects stated that :rambled words task was too simple to predict such plex skill as success in a career or as a homemaker. 3f the subjects stated that the experimenter acted :h a manner that they disbelieved her; in fact, one :t indicated she believed the experimenter because :emed honest. Many of the subjects who had been .he test was a measure of homemaking ability believed omemakers would recognize and unscramble words d to the home more rapidly than women without home- ability. An explanation favored by subjects who :eived the career version of the instructions was 1e test measured verbal ability which is part of .gence and, thus, is generally predictive of success Lreer. One additional check for experimenter effect was four-way (experimenter by sex role by fear of by task congruence) analysis of variance, using sighted means model was performed on the anagram 53 The nine subjects who had a different experimenter g the second session than they had during the first on were excluded from this analysis because of the dural requirement that there be no empty cells. As ated in Table 5, there were no significant main ts or interactions indicative of experimenter t. It would appear that while there was no experi— effect in this experiment, the subjects had con- ble doubt about the validity of the experimental lations, and, as a result, the effect of these llations was probably weakened. Scrambled Words Data The scrambled word scores were subjected to a way (sex role by fear of failure by task congruence) is of variance using the unweighted means model. nalysis was performed on the total number of cor- unscrambled anagrams and also on the interval by 1 scores. As reflected in Table 6, there were sig- t (p. <.05) main effects for sex role and task nce. Women with a career sex role orientation ed significantly better (2 = 100.449) than did ith a homemaking sex role orientation (i = 83.034). tion, subjects who were told that the task was nt with their sex role orientation performed sig— tly better (i = 100.095) than those subjects who 54 oo.oOH mmm.mmH.H oa omn.mam.v0H Hmuoe am.~m «mo.ssH.H ms mmm.mow.mm nouns CO. HHMEm MHO> flwhoo H 605.0 unmxm . .U..H N .m..m X .m.m mo. m. Hgo mHo. va.o~ H smH.ow .mxm x .u.s x .m.m mm. m” sm>o mmH. wmm.smm H mmm.amm .mxm x .o.s x .mnm Hm.H m um>o 0H0. mmv.HH H mms.HH .mxm x .m.m x .m.m ms. mam. mmH.H Hao.smm.H H Hmo.smm.H .o.s x .m.s x .m.m mm. .NNH «mm. mma.mvn H mms.mea .mxm x .o.a no. m. um>o mum. osm.com H osm.oom .mxm x a m Ha. m uw>o mHo. Hmv.sH H sme.sH .mxm x .m.m mm. mas” omm. HHo.svs H HHo.sws .o.s x ”mum mm. .Hss mam. «mm.Hom H smm.Hom .o.a x m.m mH. m. um>o so”. VNH.mm~ H sNH.mmm .m.m x .m m . m Hw>o mHH. omm.mMH H omN.mmH Hopsmaflummxm WV m a a mo.H same. was s mmm.oam.m H mmm.omm.m woamsumqoo xmma oo.s Hmm. omm. on.mmo.H H on.mmo.H mHsHHmm no “not .VHo mmm.m mmv.~om.s H mmv.mom.a oHom xmm ? mo Esm w .macmHm m mumsvm mwnwsvw oo.ooH Ham.me.H mm MHm.mom.vHH Hmuoa mm.~m os¢.oso.H Hm oms.mwm.em uouum sm.H mam. vsv.H mHe.mmm.H H mHs.mmm.H .o.a x .s.m x .m.m as. m. Ho>o New. mmv.mms H mms.mmv .o.s x .m.s mm. Hom. Hmo.H amo.m~H.H H smo.mmH.H .u.e x .m.m % mm. was. mmm. mmm.mom H www.mom .m.m x .m.m ms.m .sHo. omm.m www.mmm.m H www.mwm.t moamntmaoo smug mm.H va. smo.~ smm.omH.~ H smm.omH.~ musHHmm no name IIIWMMWHIIIIIIHmHHWIIr mam.m Noe.mmH.s H Nos.mmH.s oHom xmm mwmmmmwm .chmHm m ammwwm ma mwmmmmM mosmsnomnwm xmma Emummsmllwkwosd mmzuomunm. : III») M 56 told that the task was incongruent with their sex orientation (Y = 83.388). :tandard deviation of the scores for each group. Anagram Total—~Means and Standard Deviations Table 7 Fear of Failure Table 7 lists the mean No Fear of Failure Mean S.D. Mean S.D. Career 117.750 41.990 106.769 31.638 lent :ion Homemaker 96.400 29.355 79.462 29.082 Career 97.778 30.051 79.500 27.330 rruent Lion Homemaker 74.133 30.622 82.143 29.866 AS illustrated by Tables 8-13, these main effects elatively consistent across the intervals. HOW“ the exceptions were that sex role orientation did lieve a conventional level of significance in Interval 2 (p ‘<.07) or Interval 6 (p.< .064), :ask congruence did not achieve conventional levels lificance in Interval 5 (p1 <.077). In all three .he F ratios fell only slightly short of signifi— suggesting that the effects would probably have d conventional levels of significance with a sample. JIIIIIIIIIIIIIII oo.OOH mmm.mm ma HmN.sHm.m Hmuoe mm.sm omm.vm mm mom.mmH.m Hounm mm. HNH. 5mm. mms.mm H mms.mm .o.s x .m.m x .m.m Hm. m. um>o Hum. Hmm.s H Hmm.s .o.e x .m.m mo. m. Hm>o mmo. was.H H Hme.H .o.e x .m.m mo. m. um>o mac. smm.H H new.H .m.m x .m.m W mm.m Home. mom.m mNo.HmH H mmo.va mocmsnmcoo Hume mm.H mmH. mmo.m Hem.mm H Hem.mm musHHmm to “was ms.m tomo. mmm.m mmH.sMH H mMH.smH oHom xwm new.” . ems... .. m...” mouoom Emummfiw H Hm>kucHllm>05H hmzlwwufin. m oHnma 58 (IIIIIIIIII. oo.ooH Hms.mm mm Hmm.smm.m Hmuoa Hm.mm mmm.mm mm mms.mmm.m Houum sm.H mmm. mmH.H Hmm.mm H Hmm.~m .o.e x .m.m x .m.m we. mHv. New. mes.vm H mes.vm .o.a x .m.m om. Hmv. mum. evo.mm H meo.mm .u.e x .m.m sH. m. Hm>o msH. Ham.m H Hem.o .m.m x .m.m mo.m immo. mmm.m ome.mmH H ome.mmH mocmsumaoo Hmme mm. HNH. Hmm. HOH.HN H HOH.H~ wusHHmm no name mm.m one. Hem.m www.mmH H vmm.mNH m.Hom xmm mfimwm in... t 8%“ a .mfim mouoom Emuumgm 039 Hm>HmpsHllm>oa< mmzlmouca m anma 59 i! 00.00H mwN.mm mm NHo.OBN.m HmuOB mm.hh www.mv Nm oom.hhvtv HOHHm mv.N omo. omm.m vmm.mvH H vmm.mvH .U.B x .m.h N .m.m mv. mnv. mom. hmm.vm H 5mm.vm .U.B X .m.m mv.H va. mHh.H omv.mm H omv.mm .U.B x .m.m mm.H me. Nmm.H mNo.vm H mmo.vm .m.m x .m.m mw.m «mmo. va.v Hmm.ONN H Hmm.o~m wocwdeaOU Mmma vm.v UswNo. mNo.m mmm.vvm H mmm.vvm OHDHHmm NO Mmmm vm.m smoo. mmw.m mmN.Hmv H mmN.va OHOM xww use” . 3%... a .3” mouoom Hammond mouse Hm>umchllm>os¢ mmzlomusa 60 i 00.00H va.wN am vom.va.N HmpOB mm.mm www.mm Nm OHm.mmv.N HOHHH mo. m. H0>O omo. Hmm.H H Hmm.H .U.B N .h.m x .m.m mH. m. Hw>o MBH. omm.v H omm.v .U.B x .m.m mm. vvm. mom. wmm.mm H mmm.mm .U.B x .m.m mm. Hmm. mvm. mmN.NN H mmN.NN .m.m x .m.m 5H.v «mmo. mmv.v mmv.hHH H mmv.hHH wocmsumcoo MmdB mm. me. mHm. VBN.VN H VBN.wN deHHmm m0 Hmwm vm.m *OHo. mmo.h wmm.mmH H mmw.me mHom xmm Wm“... fit... . awwm a .m.% mmuoom ashamed HH OHQMB Hsom Hm>uwucHllm>OG< hmzlowuna 61 oo.OOH mas.em mm mmm.mms.m Hmuos mm.em men.Hm mm ems.omm.m Houum sH.H mom. H>~.H Hmm.ov H Hmm.oe .o.a x .m.m x .m.m as. «am. mom. mmm.mm H www.mm .o.e x .a.m mH.H new. mNN.H Hmm.mm H Hmm.wm .o.s x .m.m se.H HHm. omm.H mmv.om H mmv.om .m.m x .m.m mm.~ has. mo~.m mas.HOH H mas.HOH mosmsuscou Hmws Hm. mHH. Hem. mmo.H~ H amo.H~ musHHmm no news oo.s *moo. msm.s omm.ovm H omm.oem mHom xmm twmmmmww .u..... t ..mem .. .WM.HWM mwuoom Emummfim 0>Hm Hm>kusHllw>oc4 MMBIGGHSB 62 oo.ooH wnm.os mm omm.sem.m Hmuoe ms.mw Hmm.mm Hm mmH.mom.m nouns ma.H msH. Hmm.H mmn.sm H mms.sw .u.a x .m.m x .m.m HH. m. Hw>o mHH. mmm.e H mmm.e .o.e x .m.m HH.H mmm. oe~.H oeH.mv H oeH.mH .o.a x .m.m mH. m. um>o mmH. Hmm.m H Hmm.m .m.m x .m.m mm.m .soo. on.m mmm.omm H mmm.omm MWoqwsumsoo Hmme wH.H NHN. mmm.H Ham.sm H Ham.sm wusHme mo ummm m~.m Hwo. mmm.m Hmm.m~H H Hmm.wNH mHom xmm “wommmww .mHsmHm m mummwm mo wwwmwwm monoom Emummcfl me Hw>uwusHllm>os¢ mmzlmmune 63 Considering the centrality of fear of failure to formulation of this experiment, the significant find— for the main effect of fear of failure in the third :rval is both gratifying and perplexing. It is :ifying because it supports our contention that fear 'ailure is a relevant variable in predicting women's ormance. On the other hand, it is perplexing that effect was significant only during one of the six rvals. As explained in the Methodology section, the Le as a whole can be regarded as having a career Itation because the subjects indicated that having reer had significantly more importance to them than )eing married and having a family. Consequently, Lnagram total scores were re-analyzed separately for 'roup of gs who had been told that the test measured r ability and for the group of §s who had been told the test measured homemaking ability. The results ese 2 X 2 (sex role by fear of failure) unweighted analyses of variance are summarized in Tables 14 5. For the group which had received the career lotions, there was a significant (p,< .002) sex effect: The career subjects performed better .12.260) than did the homemaking subjects (2 = t). For the group which had received the homemaking Lctions this effect was virtually nonexistent 64 oo.o0H Hmo.nmm.H me oom.vmv.mm Hmuoa oe.ms mms.mmH.H we MHm.Hsm.mm uouum mm.H mvm. mom. mmm.omo.H H mmm.omo.H .m.m x .m.m «o. m. Hm>o «No. wwm.mm H www.mm oHsHHmm no such Ho.o~ tmoo. mmn.HH va.wmm.mH H va.mmm.MH mHom xmm moundvm . mumswm mmumswm mo sum w chmHm m saw: no mo Edm umoe HooHMUIIwHoom Emnmms4||w>oc4 waIOBB vH OHQMB 65 oo.OOH mno.Hmm mv vm¢.Hmm.vw HmuOB hh.Hm mmm.HHm mv mmv.HHo.Hv HOHHM Ho. m. Hw>0 moo. mom.m H wom.m .h.m x .m.m HN.m Hmo. mNo.v HHm.mmmem H HHm.mmmem wHSHHmm NO Hmmm Ho. m. Hm>o boo. vmm.m H vmm.m wHom xwm .mfimem it... . amen“ s .m.% puma mstmEmEomlleoow finmws¢nnw>os¢ analog mH mHan 66 = .007). Perhaps the task was viewed as role incon- ent by both groups of subjects when it was described a test of homemaking ability. For the homemaking tructions the main effect for fear of failure was :lose to conventional levels of significance (p < L) that it could be considered acceptable. On the emaking task, fear of failure subjects performed :er (2 = 97.089) than did subjects without fear of ,ure (Y = 79.481). If we regard the homemaking task ription as making the task incongruent for our 1e, it appears that in the incongruent condition ects with fear of failure perform better than sub- 5 without fear of failure. The significant sex role :t discovered in the original analysis of variance the anagram total score (see Table 4) appears to be 3k congruence effect. Level of Aspiration Five measures of level of aspiration and two of ' combinations were included in the data analysis. 16 summarizes the findings of a two factor (sex by fear of failure) unweighted means analysis of nce for subject's general preference for task dif— ty. There was no significant main effect or inter- n for this variable. It seems likely that the absence of significant :s for this variable was a function of the vagueness 67 00.00H [III omm.mmm mm hNN.mmm.vm HMUOB om.mm mmN.omm mm Hmm.homamm HOMHM Nw.N mHH. mvm.m mmH.mHm H mmH.me deHHmm mo numb M oHom xom hm. mmm. mmm. mvm.hmm H mvm.hmm OHDHHmm MO Hmwm Hm. vwv. flaw. mmo.th H mmo.th wHom xwm ........ . new. .. ....... MHHSOHMMHQ Mums new wosmuwmwmm pmNHHmumswwllm>oc4 hmZIoBB mH OHQMB 68 the question. This conclusion is supported by the nificant findings when subjects were asked to indi- e in which percentile they would try to have their gram task performance fall. As Table 3 indicates, aree—way analysis of variance using the unweighted 1S model produced two significant effects and a third act which approached significance. The level of per- Iance (Y = 95.844) which women with a career orien- .on said they would try for was significantly (p < higher than the level of performance indicated by women with a homemaking orientation (i = 90.483). e was a significant (p < .05) interaction between role orientation and fear of failure. Most of this at is attributable to the fear of failure subjects. ar-oriented subjects with fear of failure had a mean for" percentile of 97.778 (n = 17), while homemaking— 1ted subjects with fear of failure had a mean "try percentile of 87.229 (n = 31). A t-test reveals difference to be significant beyond the .05 level 2.2979, df = 48), using a two-tailed test of sig- ance. For the subjects without fear of failure, was virtually no difference between the two groups; for the career gs was 93.910, and mean for the aking gs was 93.736. Although not satisfying ntional levels of significance (p <.062), subjects a congruent condition tended to aspire higher than 69 2 in the incongruent condition. Table 17 lists the and standard deviation of the level of aspiration as for each group. Table 17 .evel of Aspiration--Means and Standard Deviations Fear of Failure No Fear of Failure Mean S.D. Mean S.D. Career 100.00 0.000 96.154 6.505 ruent .tion Homemaking 91.333 11.872 94.615 7.763 Career 95.556 7.265 91.667 8.349 Lgruent .tion Homemaking 83.125 22.426 92.857 11.387 Much to our surprise, a three-way analysis of nce, using an unweighted means design, on the sub- ' estimates of the percentile in which they expected performance to fall yielded only one significant LO. As Table 18 indicates, career-oriented women :ed to perform (Y = 74.509) significantly (p < .006) ' than did subjects with a homemaking orientation 7.292). Table 19 lists the mean and standard ion of the level of expectation scores for each An identical analysis of variance performed on fference between the "try for" and "expect" per— ss failed to yield any significant F ratios. There 70 0 mHv. mvo.~m H mvo.Nm .U.B x .m.m mo. m. H®>O mmo. vmm.m H vmm.m .m.m N .m.m mm. m. Hm>O va. mhv.mm H mhv.¢m OUGOSHEGOU Mmme nm.H me. «mm.H omv.mmm H omv.mmm stHme m0 Hmom om.h a.moo. mmN.m onm.mmmaH H onm.wmmaH OHOM xwm We...” . HH. .. we“ :oHuwuoomxm mo Ho>mqllw>os< mmzlmonns mH wHQma FIIIIIIZIT__________________________________—____7”” 71 [8, however, a trend approaching conventional levels of .gnificance (p.< .078) for the sex role by fear of ilure interaction. Although not significant, the dif— rence between level of aspiration and level of expec- tion tended to be greater for career subjects with fear failure (2 = 24.722) than for those without it (i = .949). For homemaking §s the trend was in the opposite fection but not of nearly as large a magnitude (fear L 22.000, no fear 2 = 24.382). Table 19 Level of Expectation--Means and Standard Deviations Fear of Failure No Fear of Failure Mean S.D. Mean S.D. Career 75.000 11.952 76.923 10.316 ’ruent ition Homemaking 67.333 10.998 66.923 13.775 Career 71.111 9.280 75.000 9.045 ngruent ition Homemaking 63.125 16.621 71.786 11.703 The Relationship Between Fear of Failure and Fear of Success Approximately six months after the scoring for 1e press imagery was completed, the same stories rescored for fear of success, using Horner's new n. Table 20 lists the mean hostile press imagery 72 and mean fear of success score for each stimulus and the standard deviations. Inspection reveals the fear of success scores to be slightly more uniform over the four stories than the hostile press imagery scores. For both scoring systems, the "kitchen" stimulus produced the most leed press imagery. Also contained in Table 20 are the 'earson product-moment coefficients of correlation between he hostile press imagery scores and the fear of success pores. These correlations are significant beyond the 001 level for all four stimuli; it appears that there 5 significant overlap between the hostile press imagery :oring system and the fear of success scoring system. Table 20 Hostile Press Imagery Scores, Fear of Failure Scores, and Correlations Hostile Press Fear of Success Imagery Score Score correlation Mean S.D. Mean S.D. tchen 1.040 1.860 2.440 1.651 .463* fice .390 1.113 1.510 1.936 .330* )ughtful )0k .710 1.589 1.890 1.964 .368* 1e .180 .753 1.670 1.625 .359* * Significant beyond .01 73 Horner's fear of success scoring system contains x components——noncontingent negative consequences, ntingent negative consequences, interpersonal engage- nt, relief, absence of instrumental activity, and sence of others. Each of these components has been signed (by Horner, 1973) a scoring weight ranging from to -2. Pearson product—moment correlations were com- ed separately for each story and for the total scores ween the numeric fear of failure scores and the numeric hr of success scores. Correlations were also computed tween the absence or presence of hostile press imagery l the absence or presence of each component of fear of cess. These data are summarized in Table 21. Exami- ion of the table reveals a relatively consistent cor- ation between hostile press imagery and the two cate- ies concerning negative consequences. The source of negative consequences, however, appears to be related :he content of the stimulus. For the "kitchen" story negative consequences appear to have been mostly ingent, while the "office" and "thoughtful look" uli appear to have elicited stories with noncontin— negative consequences. Overall, there was a 1y significant (p < .001) correlation between the ance of hostile press imagery and the appearance of :ive consequences, both contingent and noncontingent. : is, therefore, reason to suspect that Horner's new IIlI||lIIIIIIII::::f———————————""——""""“‘*“““*““““"‘*’ 74 Table 21 Correlations Between Hostile Press Imagery Scores and Components of the Fear of Success Scoring System Abs Non- Intpl. ' Abs. of Contg. Contg. Engm. Rel. intl. Others ct. :chen ERIE” .097 .512a .160 .213b .125 .105 Iple .120 .524a .167 .184 .124 .106 ice EEIt .327a .054 .152 .195 —.016 —.135 ple .298C .171 .146 .101 —.o18 —.149 ughtful 2E eric .507a .219b .134 -.086 —.035 —.090 ple .568a .186 .117 —.069 -.039 —.114 Le aric .500a .223b .101 .195 .154 .091 >le .528a .236b .106 .261c .163 .096 1. :ric .251b .352a .120 .171 .121 —.082 1e .315a .384a .137 .171 .121 -.092 29 20ntg.: Noncontingent Negative Consequences ;.: Contingent Negative Consequences L. Engm.: Interpersonal Engagement Relief Intl. Act.: Absence of Instrumental Activity of Others: Absence of Others ic: Correlation between numeric hostile press gery score and numeric fear of success score e: Correlation between the presence or absence of tile press imagery and the presence or absence of indicator of components of the fear of success ring system a = Significant beyond .001 b = Significant beyond .05 Significant beyond .01 0 ll 75 fear of success scoring system does not assess an entirely new motive, but rather, overlaps with at least one previously recognized influence on achievement behavior. Informational Data The Personal Characteristics Questionnaire com- pleted by subjects during the first session contained 27 filler items. Responses to some filler items were examined both to discover some characteristics of our sample group and to investigate possible relationships >etween the descriptive information and our experimental rariables. Let us examine some of these characteristics. 'he mean age of the subjects was 18.38 with a standard eviation of .846 and a range of 16 to 21. Ninety-five f the SS were Caucasian, while five were Negroid. The ajority of the subjects were Freshmen (72), while 19 are Sophomores, 8 were Juniors, and 1 was a Senior. tey were majoring in 27 different fields, with the ree most popular being--no preference (26), nursing 6), and education (10). Socio-economic status was timated using father's occupation (Reiss, Duncan, :t, and North, 1961), and it was discovered that the jects were from families of high socio—economic status. median socio-economic status of the subjects was in highest (10th) decile, while the mean decile of .o-economic status was 8.968 with a standard deviation 76 f 1.623. The median number of siblings was 2.0 while the ean was 2.850 with a standard deviation of 1.857. These ata indicate that our sample group tended to come from amilies slightly larger than the national average. nly one subject was an only child, 23 subjects were he oldest child, and 34 were the youngest child. The ajority of the gs (81) reported that they were in the op quarter of their high school graduating classes, with 5 of these gs being in the top 10%. Although the ajority of gs (55) had fathers who were college grad- ates, only 31 had mothers with a college education. ass than half of the subjects (48) had mothers with 1 occupation other than being a housewife. This information, along with answers to the more .titudinal questions, was subjected to both three—way ex role X fear of failure X variable) and two-way (sex 1e X variable or fear of failure X variable) chi square alyses. A number of significant (p <.05) associations :e discovered. Career subjects planned to marry at an [er age than did homemaking subjects. The fathers of jects with fear of failure tended to have more edu— ion than did the fathers of subjects without fear of Lure. For subjects with a homemaking orientation, ie with fear of failure tended to have parents who d be moderately disturbed if they left college in arison to the strong disapproval or mild approval 77 )f the parents of subjects with no fear of failure. lareer subjects were more likely than homemaking gs to lave no older sibling. Career gs were more likely than iomemaking gs to have a different religion from their >arents. Finally, subjects with a career orientation :ended to come from families with higher socio-economic status more than did subjects with a homemaking orien- :ation. Summary The sex role variable appears to be most potent f the variables examined. Significant sex role effects ere discovered for: Anagram score, level of aspiration, evel of expectation, estimated percentile on a test of omemaking ability, and difference between estimated arcentile on a test of career ability and estimated arcentile on a test of homemaking ability. Fear of ,ilure was significantly related to estimated per— rmance on a test of homemaking ability. Task con- uence was significantly related to anagram performance. ially, the sex role by fear of failure interaction )duced two significant F ratios——one for level of iration and one for the difference between the actual estimated number of anagrams completed. Table 22 contains a correlation matrix relating the variables subjected to analysis of variance luding those examined in Appendices E, F, and G) to 78 mHN.o mNN.o 00H.0I Mme.0I HhH.0l mmH.o NHH.o mmH.o vnH.OI VNH.o hvm.o FMH.o coo.H mMH.OI men.o Hmm.o Nmo.ol va.o vwm.o mvm.o mNo.o HmH.o haH.OI mmH.0I mmH.ol HmH.o mmH.o mNo.o NNH.OI mwo.OI me.0I vNo.o MMN.OI NNo.0I owo.ol mMH.OI coo.H mmv.0l va.o mmo.ol NmH.OI mwo.0I hmH.ol vVH.OI hmo.0l Hmo.o omH.o Nom.o hmH.o| MNN.ol mHH.o mmN.o Hoo.o me.o Nmo.ol HmN.o mnH.o who.o mwm.o mmv.0l coo.H HHm.o mne.0l mmv.o mHm.o mom.o mmo.0I mmH.o mmN.ol moo.o vvo.o Nvo.o mmo.0| OVH.o NmH.o vmo.0I MHo.ol mmo.0l Hwo.o va.o Hoo.ol HMN.o va.o HHm.o ooo.H 05H.ol mHm.o th.o mVH.o mmH.ol QNH.o mom.0I m moo.o vmo.ol QOH.oI mHH.0| moo.o1 mwo.ol hvo.ol who.o mOH.o HHH.o mno.o wMH.ol mmo.OI mmo.OI mno.01 05H.ol ooo.H FNN.oI mmm.OI mmo.o mmo.o moo.OI mmo.ot U moN.o mmm.o wNH.0I omH.OI mwo.ol mmN.o mHo.o chH.o NmN.ol vmo.o mmo.o om~.o ~wv.o NmH.OI wmv.o mHm.o hNN.oI ooo.H 5mm.o mmv.o mHH.ol wmm.o hHN.OI h mmm.o th.o mnm.o omN.o NMH.OI mvm.0I me.0I oom.o| mNo.o wvo.o vmm.o Ho~.o mOH.o hmo.o mmo.o omo.o QHH.ol wmo.OI mmo.o va.o mom.o mmn.o moo.o HmH.ol vnm.o mwm.o mmo.0I hmH.OI mHn.o aom.o hmm.o wVH.o mmw.OI mmo.o wmm.o mmv.o coo.H how.o nmw.o ooo.H mNN.OI ovm.OI mmo.0I Hmo.OI owH.OI vm~.ol m D ouanwm> mo mHthmcd ou uownnsm menMHum>lleHums me.oI nNH.OI MMH.o mOH.o mwo.o 0Ho.ol wNH.ol hmo.o vwo.o mvo.ot vmo.ol mwo.o m~o.o VVH.o «no.0l mmH.0I mmo.o mHH.0l mNN.OI ovm.ol coo.H HNo.OI hHo.o HVH.o mNH.o mNH.OI mOH.OI mmo.ol th.o «Ho.cl mmo.o mHo.OI hno.o HNo.0I th.o HmH.o hmo.01 mmH.o ONH.o moo.ol hMN.o mmo.ol Hmoool HNo.0I ooo.H mNH.0I GOHUMHD 4 43) mmo.ot mmo.ol mvo.o m~H.o wNH.0| owo.ol omo.o mmo.o mmo.ol mmH.ol mNm.OI mNN.o hmH.OI Hmo.o mmN.ol mON.OI wNo.ol >H~.01 owH.OI vmm.on hHo.o mNH.OI ooo.H m B D > 3 H b M A E Z O m O m fl m U Q m h 0 m 79 hom.o www.cl mvn.ol mmm.0I ham.o hom.o coo.H NHm.0I mNm.0I mvN.OI vmv.o mwm.ol NHw.OI coo.H mmm.o wvv.o mmm.ol mvn.0I mmm.on mmm.o ooo.H HNH.ol NNN.OI mmm.0l QVN.OI mvv.o HNH.0I ooo.H mmo.ol nmm.o vmv.o mmw.ot NNN.0I moo.on coo.H wmm.o 5mm.o www.01 mmm.ou mo~.o: omm.ol vmo.o mvo.o wNH.oI VVH.OI moo.o hoo.ol HNH.0I hmH.0I wmo.o mmo.o: mm~.o NNH.ol mmN.o mNN.o HmH.oI OMN.OI wmo.o me.o one-o mNo.o wmo.ol mho.ol 0Ho.o who.o NHH.c QVH.o hvo.ol hmo.o mVH.0I Hmo.o mHN.o QNN.c owH.ol mno.ol HhH.OI mmH.o mmH.OI mmH.ol HmH.o mmH.o mno.o NNH.OI omH.o NON.o hMH.oI MNN.OI mHH.o mm~.o moo.o vwo.o Nvo.o mmo.ol OVH.o NMH.o moo.o vmo.0I QOH.ol mHH.OI Noo.0I wmo.ol mo~.o NMN.o wNH.oI QMH.OI MNo.0I mwm.o mmm.o th.o hMH.0I me.OI mNo.o vmm.o th.o omN.o mvm.ol oom.ol wvo.o HoN.o mVH.oI hNH.0I MMH.o mOH.o mmo.o 0H0.0I HVH.o wNH.o mNH.OI wOH.oI mmo.OI mnH.o mmo.OI mNo.OI mvo.o mNH.o wNH.OI owo.o| 3 > D .H. m M wNw.o hww.o >v¢.0I wwm.ol mo~.ol omm.0I ooo.H mmo.o vw0.0I Nwo.o hmo.0l HmH.o NHH.o mwo.0l Hoo.o vwo.ol hvo.OI mHo.o mOH.o nmo.o wNH.OI vHo.0I omo.o vmo.o mvo.o mNH.oI VVH.0I moo.o woo.OI mmo.o coo-H who.01 omo.o mNo.oI me.o mmH.o me.OI mVH.o MHo.o Nno.o OhHuo m~o.o omo.o hmo.o mmo.o mmo.0I HNH.0I bmH.OI wmo.o maO.OI mmN.o NNH.0I wwo.0I who.0I ooo.H HNo.o mNo.0I OHH.ol VBH.0I vmo.o Nwo.0I mmo.0l mOH.o NMN.OI OHH.OI mNo.0I vwo.o mHo.0I mmo.ol mmN.o mNN.o HmH.OI om~.0I wNo.o mmH.o Nmo.o owo.o HNo.o ooo.H «no.0! wNH.o VNH.o mmm.0l HmN.o Hmo.o HHH.o vmo.o mmo.o me.o mvo.0| who.o mmH.0l omo.o mNo.o mmo.0I mho.ol 0Ho.o who.o hmo.0I mNo.0I mwo.ol vmo.OI coo.H wmh.OI hvm.o «No.0I th.o va.o mho.o mmo.o mom.o mmm.o vmo.ol HNo.OI mmm.ol NHH.o mVH.o h¢o.ol hmo.o mvH.oI Hmo.o HMH.o mmH.o OHH.oI wNH.o vmh.ol coo.H hMH.o omo.OI mno.o Hoo.ol wMH.ol om~.o moo.o HMH.0I mwo.o th.o mNN.o a D > 3 H b g A z z o A o M m :2: CO U D {1.134051 GOHHDQHHuu< Hmsuwuxm paw HwanmpsH cmwzuwm oonHOMMHQ cOHuanuuu¢ HmsuwusH COHuDQHHugfl Hmsumuxm muHsoHuuHo Hume on soHuanHuuua xODH ou GOHHSQHHuufl HHHMm on COHuanHpus unomwm on COHuanHuu< moHHusoouom sH muHsonwHo rude How oosouwmonm poNHHmuosow muHHmso whoa meprEHmexm m£# m>wHHmQ Dom GHQ so ouoom oHHpsoouwm poumEHnmm new onmm poumEHumm swmsuwm mocoHoMMHo B n so wuoom oHHucwonmm pmumEHumm huHHHns Hmoumo mo umwe n so onoom mHHpsoonm poumEHumm QOHumnommxm mo Hm>mq new GOHumuHmm< mo Hw>oH cmwzuwm mocwuomeo mmHHucwouom CH sOHumuommxm mo Hm>oH moHHucwouom CH sOHumnHmmd mo Hm>oH mfimumms< mo Honfisz pmumEHnmm new Hmsuo< sowspom oocmHoMMHo pwuonEoo mfimummcd mo wmeHumm oHHusooHom pouonEoo manumos4 mo oquHumm HMUHHoEsz whoom Hmuoa Emumms< mososumsoo xmwa oHSHHmm mo umom spomusoHHo wHom xmm MUHHHnm mcmeEmEom mo umwe m MuHHHad Hooumo mo puma n so wuoom oHHusw huHHHbfi msHMmEmEom mo Hum 80 flimUQfiJfi-JUEIH'UMHE ZOmOMmBDI>3 pcwmmq pussHusoollmm OHQMH 81 ch other. As would be expected because of the nature their determination, scores for attribution to effort, ill, luck, task difficulty, internal attribution, ternal attribution, and the difference between internal d external attribution were highly intercorrelated. ores for anagram performance, numeric estimate of agram performance, percentile estimate of anagram per— rmance, and difference between actual and estimated agram performance were highly interrelated. Significant tercorrelations were found between level of aspiration, vel of expectation, and the difference between the two. An analysis and discussion of supplementary data cluding post-performance estimates (both numerical and percentiles), estimation of career ability and of home- king ability, and of performance attribution are con— ined in Appendices G, H, and I. CHAPTER IV DISCUSSION Interpretation of this study is difficult because er 30 different dependent variables were measured and alyzed. Each of these variables could be discussed l .parately, but this procedure would resemble the pro— Lrbial description of an elephant by six blind men. 1 the other hand, a global picture of each type of indi- .dual would exclude examination of level of aspiration, :rformance attribution, and performance estimate as riables of interest in their own right. This Discussion 11 borrow from both approaches. Task Congruence Hypotheses 1 through 3 concerned the combined fect of hope of success and fear of failure on per— 'mance. None of these hypotheses were confirmed. )othesis l predicted that low fear of failure subjects ld perform significantly better in the congruent con- ion than in the incongruent condition. Although the othesis was not confirmed, the trend (t = 1.385, df = p < .173) was in the appropriate direction. This 82 83 Lnding, though not significant, is important because it 1ggests that the subjects without fear of failure were )re achievement-aroused by the task description similar - their sex role orientation than by a task description ich was more dissimilar. Most likely this effect fell ort of significance because the experimental manipu- tions were poorly accepted by the subjects. Contrary the prediction of Hypothesis 2, fear of failure sub- :ts performed significantly better in the congruent 1dition than in the incongruent condition. The inhibi— :y model of fear of failure, on which this hypothesis : based, assumes that the intensity of motivation to )id failure is a function of the incentive value of cess. It predicts that the Performance of high fear failure gs will be increasingly inhibited as the incen- e value of success increases. In this experiment it been assumed that the incentive value of success is iter in the congruent condition than in the incongruent lition. It appears that the performance of the high ' of failure subjects in the congruent condition was litated relative to their performance in the incon- nt condition. This finding is inconsistent with the bitory model of fear of failure which hypothesizes for individuals with fear of failure, when hope of ass is high, fear of failure will inhibit performance. finding raises serious doubts about the applicability 84 of the inhibitory model of fear of failure to a female sample. The results of a test of Hypothesis 3 add to :hese doubts. If fear of failure has an inhibitory effect >n performance, when the incentive value of success is 1igh, subjects with low fear of failure should perform )etter than subjects with high fear of failure. For this xperiment it was predicted that in the congruent con- ition subjects with low fear of failure would perform ignificantly better than subjects with high fear of ailure. While not significant (t = 1.063, df = 47, < .294), the trend was in the opposite direction. Jain, it appears that fear of failure had a facili— iting rather than inhibitory influence on performance. Since the fear of failure predictions were not )nfirmed, the more basic hypothesis that performance is hanced when the task description is congruent to the bject's personal definition of competency bears investi- tion. In planning the experiment it was assumed that a task would be relatively more congruent for the :eer-oriented subjects when it was described as a it of career ability and relatively more congruent ' the homemaking—oriented subjects when it was cribed as a test of homemaking ability. Analysis of responses to the following questions which were not i to assign sex role ideology suggests that this imption was justified. When asked, "How important 85 do you feel that a professional career of your own is," subjects with a career orientation reported a signifi— cantly (t = 3.75, p < .001) higher mean level of impor- tance (4.98) than did subjects with a homemaking orien— tation (4.52). When asked, "How important do you feel that marriage and a family are to you," homemaking— oriented subjects reported a significantly (t = 5.5, p < .001) higher mean level of importance (4.74) than did subjects with a career orientation (3.86). The subjects gig perform significantly better on the task when it was described as relatively more con- gruent to their personal definition of competency. On . the scrambled words task, performance was significantly (p < .014) better in the congruent condition than in the incongruent condition. Although not achieving conven- tional levels of significance, level of aspiration (p < .062) was higher when subjects served in the congruent :ondition than when they Served in the incongruent con— [ition. Subjects' estimates of the number of anagrams ompleted were significantly (p, <.Ol9) higher in the ongruent condition than in the incongruent condition. he performance predictions of subjects with a homemaking rientation for a test of homemaking ability were sig— .ficantly (p < .009) higher than those made by subjects .th a career orientation. While other variables :stimated level of performance on a test of career 86 .lity, level of expectation, post-performance percen— e estimate) did not produce F ratios acceptable at ventional leveb of significance testing, the trend the results was still the same. In short, subjects dict, estimate, and actually do perform better on ks relatively more congruent to their personal goals. The description of the experiment (that it was Lgned to detect differences between women who attend .ege and women who do not) was probably relatively 2 achievement arousing for the career—oriented sub— .s than it was for the subjects with a homemaking ntation; career-oriented subjects were significantly 2.03, p < .05) more certain that they would graduate college than subjects with a homemaking orientation. aps the scrambled words task itself was perceived as :ively more congruent by the career—oriented subjects it was by subjects with a homemaking orientation. [is was the case, the following results can be 'ded as reflecting a congruence effect: Career— ted subjects performed significantly (p. <.Oll) r than homemaking-oriented subjects, aspired sig— antly (p < .035) higher, expected to perform sig— intly (p < .006) better, and estimated that they had :med better (numerical, p < .066; percentile, p.< than homemaking-oriented subjects. It is possible difference in ability between the two groups was 87 esponsible for these results, but a chi—square test evealed no significant difference between the high chool class standings as reported by the two groups f subjects (x2 = 2.608, p < .456). Task congruence as an independent in this experi- ant was confounded. The strength of the subjects' ientification with a collegiate group was a confounding lctor produced by the explanation of the experiment. Lrther confounding was a consequence of the significantly gher (t = 2.42, p < .05) level of importance attached subjects to having a career in comparison to being rried and having a family. Because of this confounding, several congruence fects can be predicted: (1) The career task should re been relatively more congruent for all gs than the Iemaking task. (2) The performance task, in general, )uld have been more achievement arousing for the career- ented subjects than for the homemaking-oriented sub- ts. (3) The task described as congruent to the sub— t's sex role orientation should have been more Levement arousing than the task described as incon- ant. If we conceive of each of these predictions as ,ors, we would predict that the level of performance, highest to lowest, should be: (1) Career gs on er task, (2) Homemaking gs on the career task, and 1 tie between career and homemaking gs on the 88 omemaking task. This, in fact, is the order of per— ormance level for the subjects without fear of failure. hus, we tentatively conclude that task congruence has significant effect on performance. Fear of Failure None of the hypotheses concerning the effect of ar of failure on performance were confirmed. These potheses were based upon the inhibitory model of fear ’ failure. Since, on the homemaking test, subjects ith fear of failure performed better than subjects with— t fear of failure at a level near conventional levels significance (p < .051), it seems possible that fear failure had an excitatory effect on performance. is hypothesis, however, is inconsistent with the 1ding that on the career test the performance of lemaking—oriented subjects with-fear of failure (i = 133) was lower (but not significantly) than that of emaking—oriented subjects without fear of failure = 82.143). It appears that neither the inhibitory 31 of fear of failure motivation favored by McClelland Atkinson (1966) nor the excitatory model motivation >red by Birney, Burdick, and Teevan (1969) explains e results. In the Introduction it was suggested that the ntive value of success at a task is determined in by the relevance of success on the task to the 89 ndividual's personal definition of competency and in art by the implications of success on the task for ratification of the individual's interpersonal needs. : was pointed out that individuals may be disposed to ilue both components equally or may value one component >re than the other. For the high fear of failure indi— idual, the social consequences component of incentive flue, we believe, has greater weight than the competency ‘mponent. For the high n Achievement individual, we lieve, the competency component of incentive value has eater weight than the social consequences component. a establishment and maintenance of a sense of personal npetency is a primary goal for the high n Achievement iividual and a secondary goal of the high fear of Llure individual. The establishment and maintenance a comfortable interpersonal environment is a primary .1 of the high fear of failure individual and a ondary goal of the high n Achievement individual. thermore, we suggest that while the high n Achievement ividual attempts to maximize social acceptance, the 1 fear of failure individual attempts to minimize .al rejection. This view of n Achievement and fear of failure ers from the traditional conceptions chiefly in its asis upon the social consequences of behavior. lievement has been defined by McClelland, Atkinson, 90 t al., as a tendency to compare one's performance gainst a standard of excellence. Although this standard 5 excellence has theoretically been assumed to be objec- -ve, operationally it contains an element of social )mparison. In their scoring manual for the achievement »tive, McClelland, Atkinson, Clark, and Lowell state, ‘ompetition with a standard of excellence is perhaps st clear when one of the characters is engaged in com- titive activity (other than pure aggression) where nning or doing as well or better than someone else is a primary concern (1953, p. 111).“ Clearly, when the 31 of performance is defined gig 3 iii another person, a purpose of activity is to enhance feelings of com- :ency while increasing social esteem. Like n Achievement, fear of failure has not been :oretically defined as having a social comparison com— Lent, but the method by which it is assessed indirectly sures the extent to which an individual anticipates dal rejection for failing at school. Traditionally, 1 u of failure has been assessed through a measure of I anxiety; i.e., the amount of anxiety admitted by jects in a situation which has implications both for ir personal sense of competency and for their social aptance. Little attention has been devoted to the : that felt anxiety may be irrational fear. Rational :s of failure in school can be accommodated either 91 rough studying more (fighting) or lowering aspirations .ight). If an individual experiences a high level of :t anxiety, it can be assumed that either his fears , irrational (not subject to modification by studying lowering aspirations) or he is somehow constrained m taking adjustive action. For example, the college dent who continually experiences a high level of test iety is prevented from avoiding the anxiety-provoking nation by his own desire to get a college degree. a if he thinks he does not have the ability to :essfully complete his program, he may persist because wishes to please his parents or because a degree is assary for the type of employment he desires. If an .vidual has an irrational fear of failure on school :5, i.e., one not justified by lack of preparation .bility, it can be assumed that he is somehow pre— .ed from lowering his level of aspiration. It is e clear that the social consequences of lowered Vidual from taking proper adjustive action. rations most often prevent the high test anxiety i 1 I l L Thus, it appears that operationally both n I svement and fear of failure contain a soc1al com— I H son component and that this component is greater in l . . {of failure than in n Achievement. It is also ible to conclude that the assessment method for hievement is sensitive to the tendency to maximize \ xi . H 92 ial acceptance through successful competition, while test anxiety assessment method for fear of failure sensitive to the tendency to minimize social rejection failure in school. The hostile press scoring system d in this study is sensitive to the tendency of the ividual to anticipate social rejection. De Charms (1968) suggests that high fear of Lure individuals may focus on the realistic obstacles difficulties in performance as a means of minimizing Lures. We would like to expand this idea. High fear failure individuals, we believe, are characterized by )nservative strategy designed to minimize their losses iocial esteem. They avoid reasonable challenges, while :he same time they maintain their personal sense of retence through a consistent pattern of success at tasks. This pattern is suggested by the reinforce- history of high fear of failure subjects found by an and McGhee (1972). The mothers of male high 31 students who had high levels of hostile press ery tended to ignore successes and punish failures. 2 success at challenging tasks is likely to raise the >rmance expectations of socially reinforcing agents Lers, fathers, teachers, etc.), the best strategy .n individual in this situation is to attempt only on which success is likely. An alternate possi— y is to attempt tasks on which failure is virtually 93 tain so that merely attempting the task is viewed as itorious and failure is not punished. If the indi- ial is successful on an extremely difficult task, he verbally attribute his performance to luck or suggest : the task was easier than he expected while silently )ying his increased feeling of personal competence. Because the high fear of failure individual seeks ivoid losses of social esteem, he must be especially antive to the social consequences of his behavior. 1 the positive consequences of success and the nega- : conSequences of failure are greater than the nega- 2 consequences of success and the positive conse— Lces of failure, his performance will be facilitated. .he negative consequences of success and the positive equences of failure are greater than the positive equences of success and the negative consequences of ure, his performance will be inhibited. Success on a particular task can have positive a1 consequences if it confirms or raises the social am of an individual and negative social consequences : implies deviation from normative behavior. For Ile, an overwhelming victory at chess has positive Ll consequences for the individual if he is partici- .g in an international competition but has negative 1 consequences for him if his opponent is a young ive playing his first game of chess. The 94 anservative strategy adopted by fear of failure indi— Lduals allots success a positive value if it confirms :her's expectations but not if it exceeds these expec- itions because such success will increase the expec- itions others have of his performance. Failure, also, 1y have positive social consequences if it is consistent .th other's expectations and negative social consequences 5 it falls short of expectation. For example, an :ceptionally bright child is likely to be better liked ’ his peer group if he sometimes fails, but the same dividual will experience social rejection if he fails an important assignment as an adult. The performance a high fear of failure individual, therefore, will facilitated when success is expected and failure is rongly aversive. On the other hand, performance will inhibited if success is unexpected and failure con— stent with expectation. This model can explain the behavior of the fear failure subjects in this experiment. For the career- .ented subjects taking the career test, the positive :sequences of success and the negative consequences of lure outweighed the negative consequences of success the positive consequences of failure. One of the ler items in the Personal Characteristics Questionnaire ed the subjects to guess their parent's reaction 11d they drop out of college to get married. 95 ireer-oriented subjects with high fear of failure sported the highest mean level of disturbance (4.29), zile homemaking-oriented subjects with high fear of [ilure reported the lowest mean level of disturbance .16). Failure on the career task was, therefore, 'obably viewed as more socially aversive to the career— iented fear of failure subjects than it was to the memaking—oriented fear of failure subjects. Further— re, as indicated by their questionnaire answers success a career was more important to career-oriented sub- cts than it was to homemaking-oriented subjects. nally, the career-oriented subjects plan to have a reer, while the homemaking—oriented subjects do not. maintain social esteem, the career-oriented fear of Llure subjects had to perform well on the career task; .s, in fact, is what they did. The homemaking-oriented fear of failure subjects, ) accept the traditional female role, probably perceive cess at a career as interferring with their primary 1 of acquiring a husband. Success at a career is bably viewed by these subjects as more difficult than cess as a homemaker; certainly this is the view of t of our society. It seems likely that the choice a homemaking orientation by a fear of failure indi— ual (as will be explained in the section on level of iration) is part of a strategy of aiming low. These 1 1 E 5 E 96 bjects would probably view success on a test of career ility as success at an appropriately challenging task. the subjects were successful at this task, they might igine that their teachers, friends, and boyfriends 11d expect a higher level of performance than the sub— :ts felt they could easily produce. Clearly success a test of career ability had strongly aversive social sequences for the homemaking—oriented subjects with r of failure, while failure on the career test was not aversive to the homemaking-oriented subjects as it to the career—oriented subjects. It is entirely sible that the fear of failure subjects with a home— Lng orientation displayed a relative decrement in per— Iance when the task was described as a test of career .ity. When the task was described as a test of homemaking Tity, the performance of fear of failure gs was uni- %y facilitated. Failure on a test of homemaking dty could have strongly aversive social consequences these subjects. Unfortunately, the subjects were not \ l 1 d to estimate their parents' disturbance should they :to marry. It seems reasonable to assume that nearly 9f the subjects' parents would be disturbed if their lters failed to marry, since marriage is normative iomen in this culture. Ninety-three point five per— ‘of the adult women living in the United States in i 97 972 had been married at some time during their lives U.S. Bureau of the Census, 1973). Furthermore, only our out of our 100 subjects do not plan to be married 5 years from now. Failure on this task would have been versive to our subjects not only because women are (pected to succeed as homemakers but also because it )uld have implied that the subjects were unable to per- >rm well in comparison to a group of women who did not :tend college. It is, therefore, consistent with our :pectations that the performance of subjects with high :vels of fear of failure was facilitated on the homemaking sk. Preference for Task Difficulty Five measures of level of aspiration were included this study. The first of these questions asked the ojects their general preference for probability of :cess. Hypotheses 4 and 5 predicted that subjects :h low fear of failure would prefer intermediate pro- )ility of success (30-70%), while subjects with high Lr of failure would prefer extreme probability of cess (0—29% and 71-100%). Neither of these hypotheses confirmed; there was no difference between the pro- ility preferences of subjects with high and low levels fear of failure in terms of preference for intermediate extreme probabilities of success (X2 = .503, p < .5). analysis of variance examining sex role orientation 98 ; well as fear of failure (see Table 13) also failed to reduce significant effects. Perhaps the question was )0 vague to arouse the defensive operations of the fear 5 failure subjects. Another possibility is that the lestion, which is somewhat confusing, could have been Iterpreted by some subjects as asking for difficulty evel (l—PS) and as asking for probability of success ’3) by others. If this were the case, we would expect bi-modal distribution, but the combined distribution .3 a single mode at 80% Ps' It remains unclear why eSe groups were not significantly different. Perhaps l the subjects misunderstood the question. A more difficult result to explain is the mean level of 68.524. In the research literature on males, iian level of aspiration is usually about .7; this is Jivalent to a PS of .3. Median level of P5 in this idy was .80. These subjects apparently prefer tasks which success is highly likely. To the author's >wledge only in Horner's (1968) study have female )jects been asked to indicate their preference for 1k difficulty. In that experiment subjects serving the noncompetitive condition were asked to select one k from a group of seven specially selected tasks ging in difficulty from very easy to very difficult. median choice was task No. 5 (P about .29), and the S 1 was 4.63; this result is consistent with the male 99 Lndings. The Horner study was conducted at the University 5 Michigan, which has a student body quite similar to xat at M.S.U., so it seems likely that the gs in the :esent study misread the question as asking for diffi- llty level. There is, however, another possibility. )rner's study and all the other level of aspiration (LA) :udies reviewed by the author asked for difficulty 'eference in connection with a specific experimental .sk. In the present study subjects were asked to indi- te their general preference for task difficulty. Raynor 969) suggests that subjects may be highly motivated to cceed on tasks with high Ps if success on the task is rceived as instrumental to achievement of an important ture goal. Since a probability preference question antical to the one used in this study has not pre- ausly been used, it is impossible to know whether this sult is anamolous because the subjects misinterpreted : question or deviated from the norm or whether this [ding is consistent with normative behavior. Level of Aspiration After receiving the task and condition instruc- ns, subjects were asked to indicate the percentile in ch they would try to have their performance fall. otheses 6 and 7 predicted that in the congruent iition subjects with low fear of failure would have Lgnificant preference for intermediate levels of 100 spiration (30th-70th percentiles), while subjects with igh fear of failure would have a significant preference 3r extreme levels of aspiration (0-29th and 71-100th per— entiles). These hypotheses were not confirmed because Ll but five gs aspired to what was defined as an extreme evel of aspiration; the mean LA was 93.163. Hypothesis 8 us also not confirmed; on the congruent task career— ‘iented women with fear of failure did not aspire sig- ficantly higher than career-oriented women without fear failure, but the trend was in the predicted direction. ere was a significant (p <.04l) sex role by fear of ilure interaction which appeared on analysis of riance (see Table 3). Relative to the whole group, ar of failure subjects with a career orientation pired high (Y = 97.778), while fear of failure subjects :h a homemaking orientation aspired low (i = 87.229). 3 aspiration levels of subjects without fear of .1ure were intermediate (career i = 93.910, homemaking = 93.736). The difference between the levels of )iration for the fear of failure subjects was sig— ficant (t = 2.2979, df = 48) beyond the .05 level ing a two—tailed test of significance. It has con— itently been reported in the literature on males rney, Burdick, & Teevan, 1969; Atkinson & Feather, b; Heckhausen, 1967) that individuals with high fear failure tend to set their levels of aspiration either 1 l 1 101 extremely high or extremely low relative to the LA's of Low fear of failure subjects. The present results suggest :hat this finding holds true for females. Within this college population it appears that :he defensive strategy of aiming high is preferred by 'omen with a career orientation. Fear of failure sub— ects with a career orientation have a significant prefer— nce for aspiration to an extremely high level of per- ormance (X2 = 7.12, p < .01). Fear of failure subjects ith a homemaking orientation have no differential reference for defensive strategy (X2 = .03, p <.90). 1 other words, career-oriented fear of failure subjects ive a significant preference for aiming high, while )memaking—oriented fear of failure subjects have an [ual preference for high and low levels of aspiration. Long fear of failure subjects, those with a preference Ir the strategy of aiming low have a significant ten— ncy to be classified as homemaking oriented (X2 = 7.12, < .01). There was no opposite tendency for fear of ilure gs who aim high to be career oriented (X2 = .13, )<.80). In other words, those fear of failure subjects b prefer low levels of aspiration tend to have a home— king orientation, but the fear of failure gs who efer high levels of aspiration were equally divided Esex role orientatiOn. There were no significant dif— 1 bences in strategy or sex role assignment among r i subjects without fear of failure. It appears that fear f failure subjects who have a consistently low level of spiration tend to have a homemaking orientation. From common sense point of view, this finding is reasonable ecause being a homemaker is usually perceived as easier han having a career; it is also the role least sus- eptible to social criticism. It appears that selection of a career orientation ;y fear of failure subjects is independent of level of spiration. Perhaps this result is a function of the arsh criteria used for identification as a career- riented subject in this study. Remember that having a areer was significantly more important than having a irriage and a family for our gs in general. It was >und that career-oriented gs had a significant prefer- 1ce for high levels of aspiration. Had the sample been ass restricted, it would probably have been found that ear of failure gs who prefer high levels of aspiration :nd to be career oriented, while fear of failure gs who efer low levels of aspiration tend to have a homemaking ientation. While fear of failure subjects were significantly = 4.08, p < .05) more likely to be classified as ing a homemaking orientation than as having a career entation, there was no comparable trend for subjects hout fear of failure (X2 = .08, p < .80). The finding 103 that the majority of fear of failure subjects select a homemaking orientation is consistent with our theory that fear of failure is a strategy designed to minimize social rejection; women are more likely to be viewed as socially ieviant if they enter a career than if they become home- nakers. Why do some fear of failure subjects prefer the strategy of aiming high? Actually this question (and the rhole preceding discussion) assumes that individuals have .consistent preference for one defensive strategy; this ssumption has never been tested experimentally. Should his be the case, however, it is possible that accelerated arental demands for independent behavior, a variable ound by Teevan and McGhee (1972) to be significantly slated to hostile press imagery, may be responsible. Dme parents may not only have unrealistically high :pectations for their children but may also punish leir children if they do not attempt to satisfy these :pectations. These parents might punish their child >r low aspirations but not punish them if they fail to ‘hieve the unreasonable high level of aspiration. An dividual growing up under such a reinforcement system uld continuously set his level of aspiration extremely gh. Like the individual who prefers the strategy of ning low, he would seek to avoid unexpected successes tasks which constitute reasonable challenges because :h successes might raise the expectations of the .nforcing agent. 104 There are three other levels of aspiration find- gs worthy of discussion. There was a significant (p < 35) sex role effect for level of aspiration; career- iented subjects aspired higher than subjects with a nemaking orientation. As previously suggested, if a nan plans to have a career, she is in a sense setting ligh level of aspiration because of the obstacles which :iety sets in the path to her success. This explanation, )ne, may be responsible for the sex role effect. On a other hand, the scrambled words task was perceived a better test (p. <.059) when it was described as a i t of career ability than when it was described as a \ t of homemaking ability; it may have been generally 4 ceived as more related to success in a career than to cess as a homemaker. Greater ego involvement on the t of the career-oriented subjects may have been ponsible for this sex role effect. The task con- znce variable probably failed to achieve conventional als of significance (p < .062)_because the experimental .pulations were poorly accepted by the subjects. Finally, the unusually high mean level of aspir- n (93.163) requires explanation. The research liter— e on men and the results of Horner's (1968) study est that a level of aspiration of about 70-75% dif- 1ty is typical of most subjects. In this study acts were asked to indicate their level of aspiration before they had any experience with the task and without reference to a specific group. Without such specific 'nformation, perhaps the subjects guessed that they would erform as well as they had previously on standardized ests of verbal ability. Cass and Birnbaum (1972) report he mean SAT verbal score for M.S.U. students to be 550; e might expect the scores of women alone to be slightly igher. The College Entrance Examination Board (1963) ndicates that a verbal score of 550 is equivalent to blacement at the 90th percentile for all female high chool seniors. Our mean level of aspiration of 93.163, herefore, is consistent with the subjects' general level f verbal ability relative to all high school seniors. Level of Expectation Level of expectation is different from level of spiration in that it reflects the subject's realistic stimate of her performance. As such, it is an indirect ~asure of subjective probability of success. Analysis i the level of expectation data (see Table 18) produced 1y one significant effect. Career-oriented subjects pected to perform significantly (p < .006) better than emaking—oriented subjects. Reported high school 55 standing, as an indication of ability, does not m to account for this difference. It is, however, '11 possible that the career-oriented subjects had ater verbal ability than the subjects with a 106 >memaking orientation. High school class standing is function both of ability and effort. Furthermore, hools vary both in the mean ability level of their udents and in the difficulty of the material taught; "A" average in one school might be equivalent to a " average in another. It is entirely possible, in spite the fact that there were no significant differences in h school class standing between the two groups, that average verbal ability of the career-oriented sub- :ts was significantly greater than the average verbal Llity of subjects with a homemaking orientation. Another possibility is that the career-oriented >jects were more ego involved by the imaginary compari— L of their performance to that of women not attending lege. Since a t-test reveals that career—oriented jects are significantly (t = 2.03, p < .05) more tain about graduating from college than subjects with omemaking orientation, this explanation seems sup- ted. There was no significant fear of failure effect level of expectation, but there was a tendency <.l65) for the expectation levels of subjects with : of failure to be lower than those of subjects with— fear of failure. The direction of this finding is istent with at least one published study using male ects (Feather, 1965). 107 Fear of Failure vs. Fear of Success One purpose of this experiment was to compare :he hostile press scoring system for fear of failure to [orner's new (1973) scoring system for fear of success. 15 Table 21 indicates, significant (p <.001) correlations Iere found between hostile press imagery and the categories >f contingent and noncontingent negative consequences in :he Horner scoring system. Since Horner has not, as yet, )ublished this scoring system, it is possible that she ias recognized the overlap between these two measures. [orner developed her new scoring system by comparing the :tories of subjects who in competition demonstrated a iecrement in performance to those produced by gs who did .ot demonstrate a decrement. Because individuals with igh levels of fear of failure tend to perform better hen working alone than when working in competition Ryan & Lakie, 1965), this procedure undoubtedly labeled igh fear of failure subjects as having fear of success. Because the leads used in this experiment did not lclude an instance of being successful, it is impossible > mathematically compare the hostile press imagery rores to Horner's old scoring system. Simply on the sis of content, however, they appear to have much erlap. The "negative consequences because of success" tegory in the fear of success scoring system is liniscent of the hostile press of environment which 108 forms the basis of the hostile press scoring system. A major difference between the two systems is that while hostile press forces itself upon the protagonist, fear of success imagery may involve the protagonist's own reactions to her success. Fear of Failure in Women In many ways this experiment can be regarded as an exploratory study of fear of failure in women. Because many different variables were examined, it is possible to develop a typological view of the fear of failure woman in comparison to a woman without fear of failure. The interpretations which we will make below are to be regarded as very hypothetical. Since this is the first comprehensive study of fear of failure in women, many replication studies will be required to establish the validity of these conclusions. The woman with fear of failure appears to have Less confidence in her abilities than a woman without Fear of failure. Her level of expectation has a non— iignificant (p <.l65) tendency to be lower than that of ubjects without fear of failure. Her ability estimates re lower than those of subjects without fear of failure 3th on a test of career ability (p <.lO9) and on a est of homemaking ability (p <.024). Her estimates 7 relative performance are significantly lower (p < 20) than those of subjects without fear of failure in 109 spite of the fact that her performance tends (p < .154) :o be better than that of the subjects without fear of failure. It is a mistake to lump all fear of failure women into one category because there are numerous differences oetween those with a career orientation and those with a homemaking orientation. Career-oriented women with fear of failure set their level of aspiration extremely high. They over-estimate the performance of others and under— estimate the relative level of their own performance. While they believe their ability to succeed at a career is above the average, they believe their ability to nanage marriage and having a family is below the average. )n the other hand, fear of failure subjects with a home— making orientation set their level of aspiration extremely .ow. They estimate their performance, both numerically 1nd relative to others, to be low. Although they believe .heir ability to succeed as a homemaker is slightly reater than their ability to succeed at a career, they elieve they are above average on both abilities. Examination of these data suggests that fear of iilure subjects with a homemaking orientation have a enerally low opinion of their abilities and behave in manner designed to avoid loss of social esteem through ilure. These subjects, we believe, would be labeled having high fear of success by Horner's old (1968) llO scoring system. Success on a role inappropriate task would probably be viewed by them as unlikely and not worth the risk of loss of social esteem. Although these women will most likely reduce threats to their social esteem by following the conventional female role model, they will never be as secure as their counterparts who do not possess high levels of fear of failure. The career—oriented woman with fear of failure aspires to the role which is believed by the culture to be more difficult, but she feels that her ability to perform the traditionally female role activities is less :han average. Her level of aspiration is extremely high :elative to the aspiration levels of other women. She >erceives others as much more capable than they actually Lre, and, as a result, she probably experiences constant .nxiety over the possibility of 50cial rejection and ,bout her personal competence. Her high aspirations will orce her to attempt tasks on which her chances of being uccessful are low. Because she misperceives the per— ormance of others, she will strive for a standard of erfection which will make those around her uncomfortable. inally, if she fails at her career, she will be forced etc a role for which she feels inadequate, as a home~ sker. While all fear of failure women appear to lack >nfidence or feel inadequate, the two sex role lll orientation groups appear to have adjusted to these feelings differently. It appears that those with a homemaking orientation view themselves as somewhat deficient in ability and, as a result, have chosen to gratify their achievement needs through social accomplish- ments. Those with a career orientation, on the other hand, appear to feel inadequate to fill the conventional female role but believe themselves to be capable in the traditionally masculine realm of the career. As we have previously hypothesized, it seems likely that they come from homes where the traditional female role is not valued as much as the traditional male role. These neurotic characteristics are descriptive >nly of women with fear of failure. It should be pointed >ut that less than half of the career-oriented subjects [Ere classified as having fear of failure. Those with- )ut fear of failure appeared to feel confident of their .bility to succeed both in a career and as a homemaker. Fear of Failure: A Theoretical Conception Throughout this discussion it has been asserted hat fear of failure is a strategy designed to minimize ocial rejection. This assertion is a major deviation rom the previous conceptions which have viewed it as a 3tive. As a motive, fear of failure has been conceived 5 either constantly inhibiting performance (Atkinson & 112 Feather, 1966) or facilitating performance up to some maximum (Birney, Burdick, & Teevan, 1969). Kukla (1972) suggests that the experimental results found in studies of fear of failure are artifacts produced by a simple difference in perceived ability level. None of these hypotheses seem adequate to account for the findings of the present study. The strategy conception presented here suggests that fear of failure sometimes facilitates and sometimes inhibits performance. The function of the fear of failure strategy is to avoid social rejection while producing a low but consistent level of performance to naintain a personal sense of competence. Reasonable :hallenges as avoided by individuals utilizing this strategy because, as Mettee (1971) suggests, success at Fhallenging tasks is likely to raise the level of per- iormance expected by others. Extremely difficult tasks Lay be attempted as a means of demonstrating that the ndividual is not a coward. Failure at such tasks is iikely to be acceptable to others, and success can be iewed by others as a turn of luck which does not reflect rue ability while the individual secretly enjoys ncreased feelings of competency. These individuals hould not persist in their attempts to solve an “easy" ask which proves to be difficult because they expect to a rejected for their continued failure. On the other hand, they may persist at a difficult task because success on this task can be owned. Because the rejection, the fear the self-confidence 113 secretly enjoyed but publicly dis- strategy aims at avoiding social of failure subject will never have of the individual unburdened by this strategy. He cannot allow himself to attempt tasks of medium difficulty because of the risk of failing at what will appear to others as an appropriate challenge. Instead, he confines himself to easy successes which do little to increase his self-esteem. When he succeeds at a very difficult task, it is unclear to him whether his ability was responsible for his success. What evidence is there to support this contention? Po begin with, if high fear of failure individuals seek :o avoid social rejection, they should perform better Ihen working alone than when working in competition. tyan and Lakie (1965) found this to be the case. In Iorner's (1968) experiment, male subjects with high esultant achievement motivation performed significantly etter in competition than when working alone, while men ith low resultant achievement motivation performed sig— Lficantly better when working alone than when working 1 competition. Similarly, high fear of failure indi- .duals should perform better under success feedback than der failure feedback; this finding has been well cumented (Weiner, 1965; Schrauger & Rosenberg, 1970; 114 Karabenick & Marshall, 1974). They should exert more effort at succeeding on easy tasks than subjects without fear of failure, and they should be more concerned about obeying the letter of the law (Heckhausen, 1967). When presented with a difficult task, they should do better when the significance of failure is minimized than when it is made to seem important (Sarason, 1961). Implications and Directions for Future Research Clearly this study indicates that fear of failure is a relevant and important variable in the study of achievement behavior in women. Furthermore, it suggests that any study of the achievement behavior of women must take into account both the goals of the subject and the relevance of the experimental task to achievement of :hese goals. The supposedly inconsistent results in >revious studies of achievement motivation in women most .ikely resulted from the experimenters' failure to recog- Lize: (1) That success in a career and in leadership ctivities is an important goal only for some women; 2) That women who accept the conventional female role ttempt to achieve excellence relative to that goal; 3) That failure in some situations can produce social awards; and (4) That normative goals differ not only 'om subculture to subculture but also within the same bculture across time. 115 As a basically exploratory study, this disser- tation is suggestive of many lines of research. The hypothesis that fear of failure is a strategy designed to minimize social rejection needs to be tested more directly. We have suggested that behavior which appears to be motivated by fear of success is actually mediated by the choice of goals incompatible with the success and/or by fear of failure; this hypothesis needs to be tested. A developmental study of career-oriented sub- jects with high fear of failure is needed. If, as we have hypothesized, these women grew up in families where the traditional female role was belittled, it is clear that some of the thrust of the women's liberation move— ‘ment is misplaced. Women and men should be free to choose ‘their place of work (in or out of the home) out of a sense of positive interest rather than a feeling that the alternate is undesirable. Mr This study was designed to demonstrate that the achievement behavior of women can be explained by the expectancy—value theory of achievement motivation with— )ut recourse to the concept of motive to avoid success. ’redictions based upon the inhibitory model of motive .o avoid failure and on previous studies relating ego nvolvement to performance were made about performance, evel of aspiration, and level of expectation. 116 On the basis of their answers to three questions, female volunteers from introductory psychology classes were classified as having a career orientation, a home— making orientation, or as being undecided about their sex role orientation. The undecided group was eliminated. Stories told to four relatively neutral sentences were scored for hostile press imagery. A median break clas- sified subjects with any hostile press imagery as high fear of failure subjects. At a second session the 100 remaining subjects, divided into four groups on the basis of their sex role orientation and fear of failure clas- sification, were administered a scrambled words test. Half of the subjects within each group were told that the task was a measure of homemaking ability, and half were ‘told that the task was a measure of career ability. iBefore attempting the task the subjects were asked to indicate their levels of aspiration and expectation. The level of aspiration, level of expectation, erformance, and post-performance estimates of career— riented subjects were significantly higher than those f subjects with a homemaking orientation. The task ongruence variable produced significant F ratios for erformance and numerical estimate of performance. 1 Although the fear of failure subjects had an >verall nonsignificant tendency to perform better than subjects without fear of failure, their post—performance 117 percentile estimates had a significant tendency to be lower than those of subjects without fear of failure. When the task was described as a test of homemaking ability, subjects with fear of failure performed better than subjects without fear of failure (p < .051). Career—oriented subjects with fear of failure set their level of aspiration high and appeared to under- estimate their own performance while over-estimating the performance of others. While they ability to succeed at a Career was average, they estimated that their homemaker was below average. Fear with a homemaking orientation, who estimated that their well above the ability to be a of failure subjects were significantly more numerous than those with a career orientation, set their level of aspiration extremely low and estimated that their performance was extremely low. Relative to the other subjects, their estimates of ability on tests of career and homemaking ability were low, but there was little difference between their two ability predictions. It is hypothesized that incentive value has both a competence and a social consequences component and that the relative weight of the social consequences component is greater for the fear of failure individual than for the individual without fear of failure. Fear of failure is conceived of as a strategy designed to minimize social rejection while maintaining a sense of personal competency. APPENDICES APPENDIX A CUE INTERPRETATIONS APPENDIX A CUE INTERPRETATIONS ______.___._______ Instructions The following pages contain a series of verbal leads or cues. Your task will be to tell a story that is suggested to you by each cue. Fry to imagine what is going on in each. Then tell what the situation Ls, what led up to the situation, what the people are thinking and Feeling, and what they will do. In other words, write as complete a story as you can -- a story ith plot and characters. You will have five minutes to write each story. Write your first mpressions and work rapidly. I will keep time and tell you when it s time to finish your story and to get ready for the next cue. There are no right or wrong stories or kinds of stories, so you ay feel free to write whatever story is suggested to you when you look : a cue. Spelling, punctuation, and grammar are not important. What 3 important is to write out as fully and as quickly as possible the :ory that comes into your mind as you imagine what is going on in LCh cue. Notice that there will be one page for writing each story. If u need more space for writing any story, use the reverse side of e Page. Do not turn or go on to the next page until I tell you to so. Do not write your name or any other identifying marks on any the pages. 118 119 Use these questions as a guide in writing your story: 1. What is happening? Who are the persons? 2. What has led up to this situation? That is, what has happened in the past? 3. What is being thought? What is wanted? By whom? 4. What will happen? What will be done? A WOMAN IS WORKING IN AN OFFICE. 120 Use these questions as a guide in writing your story: 1. What is happening? Who are the persons? 2. What has led up to this situation? That is, what has happened in the past? 3. What is being thought? What is wanted? By whom? 4. What will happen? What will be done? A WOMAN HAS A SMILE ON HER FACE. 121 Use these questions as a guide in writing your story: 1. What is happening? Who are the persons? 2. What has led up to this situation? That is, what has happened in the past? 3. What is being thought? What is wanted? By whom? 4. What will happen? What will be done? A WOMAN HAS A THOUGHTFUL LOOK ON HER FACE. 122 Use these questions as a guide in writing your story: 1. What is happening? Who are the persons? 2. What has led up to this situation? That is, what has happened in the past? 3. What is being thought? What is wanted? By whom? 4. What will happen? What will be done? A WOMAN IS WORKING IN HER KITCHEN. APPENDIX B PERSONAL CHARACTERISTICS QUESTIONNAIRE 10. ll. . Age . In what country did you grow up? APPENDIX B PERSONAL CHARACTERISTICS QUESTIONNAIRE Name Year in school Major At what age did you decide that you wanted to attend college? Do you plan to graduate? a. Definitely yes b. Probably yes c. Undecided d. Probably no e. Definitely no How important do you feel that a professional career of your own is? a. Important b. Probably important c. Undecided d. Probably unimportant e. Unimportant Race Population of the general community in which you grew up. a. Over 1 million I). 500,000 — 1,000,000 c. 100,000 - 500,000 d. 25,000 — 100,000 e. Under 25,000 Marital status a. Single b. Engaged c. Married d. Separated or Divorced e. Widow 123 12. 13 14. 15. l6. 17. When do you expect to get married? 124 when you had expected to marry. a. b. c. d. e. Age 18 or before Age 18-21 Age 21-26 Age 26-30 Age 30 or older If you could have only a career or only marriage, which do you think you would choose? a. b. c. d. e. Definitely career without marriage Would probably prefer career rather than marriage Undecided Would probably prefer marriage without career Definitely marriage without career Children a. b. c. d. e. None 1 2 3 4 or more Would you want to work under the following conditions? a. One child of school age, husband's salary adequate 1. Definitely yes . Probably yes . Undecided . Probably not . Definitely not U'oJ-\U)N . Two or more children of school age, husband's salary adequate l. Definitely yes 2. Probably yes 3. Undecided 4. Probably not 5. Definitely not In dorm . With roommate(s) in apartment With husband . With parent(s) . In sorority or communal house ._Alone not married or engaged, do you Have a steady boyfriend . Date several different men Date sometimes Seldom go out with men except in a group Prefer not to date men If already married, answer for 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 125 How important do you feel that marriage and a family are to you? 3. Important b. Probably important c. Undecided d. Probably unimportant e. Unimportant Father's occupation Mother's occupation Father ' 3 education Mother's education How do you think your parents would feel if you stopped going to college (temporarily or permanently) in order to get married? a. Extremely disappointed, worried, or disapproving b Somewhat disappointed, worried, or disapproving c. No effect d. Somewhat relieved, pleased, approving e. Extremely relieved, pleased, approving Number of brothers and sisters Position in family a. Only child b. Oldest c. Middle child d Youngest e. Other Sex of next older sibling I a. None I b. Male c. Female Sex of next younger sibling a. None b. Male c. Female How many of your brothers and sisters have gone, are going, or plan to go to college? I was most strongly urged to go to college by a. My mother b. My father c. Sisters or brothers d. Other relatives e. Teachers f. Friends 3. No one 126 30. Fifteen years from now would you like to be a. A housewife with no children b. A housewife with one or more children c. An unmarried career woman d. A married career woman without children e. A married career woman with children f. Other 31. Parents' religion 32. Your religion 33 What effect did your religious education have on your choice to attend college? a. Extremely strong influence b. Moderately strong influence c. Some influence d. Little influence e. No influence at all 34. High school class standing 3. Top 10% b. Top quartile c. Second quartile d. Third quartile e. Fourth quartile 35. Assume that you are trained for the occupation of your choice, that you will marry and have children, and that your husband will earn enough so that you will never have to work unless you want to. Under these conditions, which of the following would you prefer? . To participate in clubs or volunteer work . To spend time on hobbies, sports, or other activities To work part—time in your chosen occupation To work full—time in your chosen occupation To concentrate on home and family . Other Humancrm P80ple differ in terms of how difficult they like tasks to be, 1 Whether academic, sports, hobby tasks, etc. Sometimes people thinc 0f task difficulty in terms of their estimated probability ofhsuciess. In ganeral, how difficult do you like tasks to be, that :5, g :t k: your general estimated probability of success on the sin 5 o as YOU most prefer? (Circle the "I" mark closest to your choice). 36 I I I I I I I I I___I____I..__.____ 0 10 20 30 4o 50 60 7o 80 90S 12:55 Impossible Z of probability of success “C Certain 127 37. Do you have any objection to serving in the second session of this experiment? Yes No 38. Your telephone number APPENDIX C SCRAMBLED WORDS TEST APPENDIX C DO NOT OPEN THE TEST BOOKLET UNTIL YOU ARE TOLD TO DO SO! On the following pages you will This is a scrambled words test. find four—letter words which have had the letters scrambled around. Your task is to unscramble the letters so that they form a word. For example, if the letters OESH were given, they could be rearranged to form the word SHOE. Some of the letters will form more than one word, but you are to write only one word on the line to the right of the letters. For example, if the letters STOL were given, you could write either LOTS 25 SLOT, but not both. No slang, foreign words, proper nouns, or abbreviations are allowed. This is a timed test. I will tell you when to start. When I say begin, turn the page and start working. When I say to stop, put your pencils down. There are six pages. Each page is timed separately. For each page, I will tell you when to start and when to stop. be able to do all of the words in the time allowed, but do as many as you can. You may do the words in any order; there is no penalty for skipping around as long as you do not turn the page. STOP! Wait until you are told to begin. ____~___________ ________________ code score _____.____.__‘____ group 128 You will not i; 129 How well will you try to do on this test? I I_____I____-I_____I_____I I I I I I 0 10 20 30 40 50 60 7O 80 90 100 CIRCLE THE "I" MARKER CLOSEST TO THE PERCENTILE IN WHICH YOU WILL TRY TO HAVE YOUR PERFORMANCE FALL. How well do you think you will do on this test? CIRCLE THE "I" MARKER CLOSEST TO THE PERCENTILE IN WHICH YOU EXPECT TO HAVE YOUR PERFORMANCE FALL. STOP! Wait until you are told to begin. 10. ll. 12. l3. 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. okob tspo . ayts . esnw rcta , kcra . msee rgud kema adys iksd cied rbmi oras aplc inrg yzla mlpu oesr lpto acer amro ogta acto code 130 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. plac eesa eidr choe aehv adie flte ahtt oruy erat aepl oodr npta plas kwla oopl omse nifd hcri ofru desu aedm Stte rahd tjsu 10. ll. 11. 13. 14. 15. l6. 17. 18. 19. 21. 22. 23. 24. 25. . letl lrea . abne . refi . scpu . rokc . orpd . psat engo elub lilp oorm aedt imte alst ubst tlod aesc . oebn ahss rhia dinw tgea leka code 131 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 26. 37. 33. 39. 4o. 41. 42. 43. 44. as. 46. 47. 48. 49. 50. lawl gihh enpo isph yamn kdco efca nwdo ofro dwna ahsh ezro psil eaht deha ande olbt adry kcbu cfat alcf nael efle alfo cikt l. 2. 3. 4. 10. ll. 12. 13. 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. ikkc eitr octl luct . eitl . uhal . nipt . enli . obss iest nadh nrho aemf sifh edhs clko eppi acsn eipn sogd ivle lwfo omrf sunp loeh code 132 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. rean kecd alir efte tase gueh ekli tafe knta bceu pmla ihts aefl olca nyra srie hcae isck wlob sewt nbed gsno lkat kjac gkni 1‘ 10. ll. 12. l3. 14. 15. 16. 17. 18. 19. 20. 21. eman . indm . ysea . sebt . dato . ivge . awns . abht daiv pcea ntiy dlna rcno uohr rutn obnr stos iptr trse lwka cpka . ogse htac . tras code 133 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. loen reev aded enke ohfo tdsu mceo alht tnxe rhee lafl nolg adel elfl krco reva reow sbta idsa gepa injo aekb itrd gsle oelp 10. ll. 12. l3. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. dege etem . kpca 08110 . netw . lcoo . pjum nsda . aflp emas dgoo neap nrga emta ncoe aetd eaar wlle rdbi frta abto ekyo L‘Wan lohd dero code WWW k 1 134 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. edhe suht rsfu ihed raeb ciem ulcb afgl blla lasi atsm stoh ikdn elsa glas uppm buet eipp aild mboc nggo sneo pmro msae osgn 1. etsn __*__________._.___ 2, ukdc ________________._. 3. wnas ________,__________ 4. mibl _____.________——-—~ 5. adrc __________,_.______ 6. card 7. ogld 8. kclu 9. hsac 10. rtya ll. ovel 12. ptea 13. docr l4. stca 15. Rose ______,_fl__._.____. 16. icen _________________—- 17. rwge ____________,______ 18. mide —— l9. illb ____________———~———— 20. uorf __*-________.,__1___ 21. iecv _~_______~———————- 22, nwot ___.____——————*~*“ 23. noil __._1__—~———-——“ 24. lswo ____~__-—-———-r‘-"‘ 25. lide ____*—-—~rr“’“" ~".______,..u code 135 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. olto sfta asef anli vihe ngto nfei aybb nswa shbu artp stpa aysh otls itls enct rion tswo mahs lfie almb olis sspa gnha osli APPENDIX D DATA QUESTIONNAIRE l. 2. 3. APPENDIX D DATA QUESTIONNAIRE FORM C Name Approximately how many anagrams do you think you unscrambled? In what percentile do you think your performance fell on this test? I I I I I I I I I I I 0 10 20 30 4O 50 60 70 80 90 100 Below Average Average Above Average . What percentage of your performance do you attribute to: Effort Luck Skill Task Difficulty . How good a test of career ability do you think this task was? a. Excellent b. Good c. Fair d. Poor . Estimate your performance in percentiles on a test predicting success in a career. I I I I I I I I I I I O 10 20 3O 40 50 6O 70 80 90 100 No Talent Average Brillant Potential Estimate your performance in percentiles on a test predicting success as a homemaker. I I I I 1 I I I I I I 0 10 20 30 40 50 60 70 80 90 100 No Talent Average Brillam; Potential 93m 136 137 Page 2 PLEASE DO NOT RETURN TO THE PREVIOUS PAGE 1. What do you think this experiment was about? 2. Did you believe the experimenter when she told you that the scrambled words test correlated with success in a career? Yes No Somewhat Why? IF YOU WOULD LIKE TO RECEIVE A LETTER EXPLAINING THE PURPgigKAIg: $31:ng OF THIS STUDY, PLEASE WRITE YOUR NAME AND ADDRESS ON THE SHEET OF PAPER. APPENDIX E SESSION I INSTRUCTIONS APPENDIX E Session I Instructions "Thank you for coming today. My name is , and I am one of the experimenters for this study. As you probably know, psychologists have recently become more interested in the psychology of women. In this experiment we are trying to find differences in per- sonal characteristics between women who attend college and women who do not. The results produced by this group will be compared to those produced by women of the same age who do not attend college. I'd like us to start by having you read the instructions on the booklet entitled, 'Cue Interpretations.'" After all the S's had read the instructions the E said, "As indicated in the instructions, I will time you while you write the stories. I will tell you when you have a minute left. At the end Of that minute, I'll say 'finish up,‘ and if you are not done by then, just finish the sentence you are writing. Do not begin the next story until I tell you to go ahead. You will have five minutes for each story. Ready, begin." After the S's had completed all four stories, E said, "For the next part of the experiment there is no time limit. Just answer the questions to the best of 138 139 your ability. Some of you will be asked to participate in the second part of this experiment, and for this reason I would like you to indicate on the last page if you would not be willing to participate in a second session for additional credit. When you are finished, please bring your booklets up to me. You can go ahead." APPENDIX F SCRAMBLED WORD INSTRUCTIONS APPENDIX F SCRAMBLED WORD INSTRUCTIONS "Thank you for returning for the second session. Today we would like to get some additional data from you which will help us in our comparison of college and non—college women. The test which I would like you to take discriminates between women who will be homemakers and women who will have careers. The test consists of a series of scrambled four letter words, 300 in all, and your task will be to unscramble as many as you can within the time limit.” 140 APPENDIX G POST-PERFORMANCE ESTIMATES APPENDIX G Post-Performance Estimates After completing the anagram task, subjects were asked to estimate the number of anagrams that they had completed and the percentile in which they thought their performance had fallen. One might expect that the find- ings on these two measures would be identical. While they were for the most part, there was one surprising difference. Both measures were subjected to a three-way analysis of variance, using the unweighted means model. As Table 23 indicates, the numerical estimates of subjects in the congruent condition (X = 98.129) were signifi- cantly (p. <.Ol9) higher than those of subjects in the incongruent condition (X = 75.861). This effect was not significant (p < .202) for the percentile estimates (see Table 24), but the direction was still the same. For both types of estimates, the sex role effect approached significance (p < .066 for numerical, p <.053 for per— centile) with the career-oriented subjects' estimates being higher (numerical SE = 95.86, percentile x = 52.329) than thOSe of the homemaking-oriented subjects (numerical i = 78.305, percentile X = 45.839). It is when we examine the fear of failure data that a rather perplexing contra- diction becomes apparent. When subjects estimated their performance in percentiles, the estimates of subJeCts 141 142 oo.oeH OHN.mmo.m mm mmm.mom.emH Hmuoe om.ew omm.sHm.H mm mNH.OHe.emH uonnm mm. m. Hm>o mmm. meo.mmm H meo.mmm .o.a x .m.m x .m.m mm. m. um>o smm. mam.emm H mam.emo .o.s x .m.m mo. m. nw>o mmo. Hem.me H Ham.me .o.e x .m.m eo.e rave. omo.e Hom.mvm.s H Hom.mem.s .m.m x .m.m em.m rmHo. oms.m mme.mmm.OH H Noe.mmm.eH moansumcoo rnms Hm.H emm. emm.H mmm.mmm.m H mmm.mmm.m musHHmm no news me.m ewe. Hee.m mmm.mmm.¢ H mwm.mem.m mHom xmm new.” ,. new. a we” oumfifiumm Emummsd HmOHHmESZIlm>os¢ hmzlwmuna MN magma oo.OOH www.mmm hm www.mmvthm HMHOB mn.mm omm.mmm om mmv.mmm.NN HOHHM mn.m HOH. hmh.N wmv.von H mmv.v0h .U.B x .m.m x .m.m mn.H 55H. wmm.H www.mnv H amw.m>v .U.B N .m.m mo. m. Hw>o mmo. mmv.m H mwv.w .U.B x .m.m m wv.H HHN. mmm.H mom.mov H mom.mov .m.m N .m.m vm.H NON. mmm.H Hvo.mwv H Hvo.mmv wocmdumcoo Mmme vm.m «0N0. mmw.m NNh.hmw.H H www.5mv.H deHHmm m0 Hmwm wm.m mmo. wvm.m mmm.mmm H mmm.mwm mHom xww wax ,. new. 2 HM...“ wumfiwumm Ewumwfim mHHunwoummlamSofim hmzlwwHaH. «N wHflmB 144 with no fear of failure (X = 53.009) were significantly higher (p < .020) than those of subjects with fear of failure (X = 45.160). On the numerical estimates, how- ever, the direction was reversed (fear X = 92.347, no fear X = 81.643). Although not significant (p < .254), there was a tendency for subjects with fear of failure to make higher numerical estimates than did subjects without fear of failure. Actually, there was a significant (p < .047) Sex role by fear of failure interaction for the numerical data. This interaction effect is chiefly attributable to the significant (p < .001, t = 4.8778 with 44 df) difference between the fear of failure subjects with differing sex role orientations. While the mean numeri- cal estimate of fear of failure subjects with a career orientation was 110.444 (n = 16), that of the fear subjects with a homemaking orientation was 74.250 (n = 28). There was very little difference between the two groups of sub— jects with no fear of failure; the mean for career- oriented gs was 80.927 (n = 23), while the mean for homemaking—oriented gs was 82.360 (n = 27)- The data were also analyzed to examine the accuracy of the post-performance numerical estimates. A three-way analysis of variance using the unweighted mean mOdel was performed on the difference between §'S actual performance and her numerical estimate of her performance. Table 25 reveals that there was a 145 mm 0Hflma {if oe.e0H msH.Hme.H Ne mme.swm.eeH Heuoe ee.mm mee.sve.H mm om~.mee.mm uonnm mm.m use. eHm.m mem.mmm.m H mmm.mem.m .o.s x .m.m x .m.m me. m. nm>o ewe. Hom.om H Hom.em .o.s x .m.m mm.H New. sNN.H mmm.wmm.H H em~.emm.H .o.s x .m.m mm.m reme. smH.m mmm.eee.m H mm~.eee.m .m.m x .m.m eH.H mam. HHH.H mHs.eeH.H H eHs.emH.H monmsueaoo rams mo. m. Hm:6 emo. mme.me H mme.mm muaHHmm no news 00. m. uo>o woo. msm.v H mnw.e oHom xom new.” . an“. .. HM...“ mumEHumm monMEHOMHmmlumom HmoHuwssz 9.3 one wusmfiuomumm Emummcd consumm wocwnwmeollm>oc¢ hwzlmonca 146 significant (p < .026) sex role by fear of failure interaction. Career-oriented subjects with fear of failure and homemaking-oriented subjects without fear of failure tended to overestimate their performance, while homemaking-oriented subjects with fear of failure and career-oriented subjects without fear of failure tended to underestimate their performance. The author was able to find only one study (Ziskin, 1966) which included estimation of performance as a dependent variable. There have been many studies examining the Zeigarnik effect in connection with fear of failure but apparently none examining either accuracy of judgment or perception of the performance of other subjects. In Fear of Failure (1969) Birney, Burdick, and Teevan suggest that high fear of failure individuals may overestimate performance as a means of defending against failure. Ziskin (1966), however, found that subjects with low test anxiety rated their performance as signifi- cantly higher than did subjects with high levels of test anxiety. Our conception that fear of failure is a defen- sive strategy designed to minimize social rejection pre- dicts that fear of failure subjects who habitually aim low should set their performance estimates low. In this fashion they can avoid being ridiculed for bragging about Poor performance. On the other hand, fear of failure subjects who habitually aim high should set their 147 estimates high. This strategy allows the individual to avoid rejection by making others believe that she has made her best attempt. If performance is not as good as the subject has indicated, she can suggest that the task was impossible. The data are only partially consistent with this hypothesis. For the numerical estimates there was a significant (p ‘<.001) difference between the estimates of career-oriented subjects with fear of failure (X = 110.444) and those of homemaking-oriented subjects with fear of failure (X = 74.250). There was virtually no dif- ference between the estimates of subjects without fear of failure. We can conclude that in an ambiguous situation, in which the individual does not know how others will perform, a fear of failure individual will follow her characteristic pattern of estimating high or estimating low. When the subjects were asked to estimate their performance relative to others (in percentiles), this pattern did not hold up. Instead, the estimates of fear of failure individuals (X = 45.160) were significantly (p < .020) lower than those of subjects without fear of failure (X = 53.009). Perhaps the fear of failure individual has a generally low estimate of her abilities relative to other's but has a characteristic extreme pattern of self—evaluation in an ambiguous situation. 148 This hypothesis is supported by the similar extreme patterns in the level of aspiration and numerical estimate data and by the poor self—confidence expressed in the level of expectation and in the post-performance percentile estimate data. Except for the behavior of career-oriented sub- jects with fear of failure, it could be said that fear of failure subjects had generally low estimates of their abilities. Relative to the other groups, homemaking— oriented gs with fear of failure had the lowest mean per— formance estimates for both the numerical and the per- centile data. The mean numerical estimate of career- oriented subjects with fear of failure, on the other hand, was nearly one standard deviation above the means of the other groups. The mean percentile estimate of this group, however, was second lowest. It appears that the career-oriented women with fear of failure overestimated the performance of the other subjects. Although they estimated that their performance fell, on the average, in the 46.319th percentile, they estimated that they had completed 110.444 words. The percentile equivalent of this number of words for this sample is placement in the 74th percentile, nearly 28 percentile points higher than their percentile estimate. To complete our discussion of performance esti— mates, we must examine the effects of sex role and task 149 congruence. Although failing to achieve conventional levels of significance, the estimates of career-oriented subjects were higher than those of subjects with a home- making orientation for both measures. This finding is consistent with the fact that the performance of career— oriented gs was significantly higher than that of subjects with a homemaking orientation. The performance estimates of subjects in the congruent condition were higher than those of subjects in the incongruent condition, but this effect achieved conventional levels of significance only for the numerical data. This finding is also consistent with the performance data. APPENDIX H PREDICTION OF ABILITY ON TESTS OF CAREER AND HOMEMAKING ABILITY APPENDIX H Prediction of Ability on Tests of Career and Homemaking Ability In the questionnaire administered after completion of the Scrambled Words Test, subjects were asked to pre- dict their relative level of performance on a test of career ability and on a test of homemaking ability. Analysis of responses to the career question failed to produce any significant F ratios (see Tables 26 and 27). There was, however, a trend (p < .078) for career-oriented subjects to predict a higher level of performance (X = 67.073) than did homemaking-oriented subjects (X = 61.156). Since sex role orientation assignment was based, to a great extent, upon subjects' commitment to a career orientation, the lack of significance is some- what problematical. The restricted range of our sample is probably responsible for this lack of significance. As previously noted, all the subjects in this sample can 3e regarded as more career oriented than homemaking )riented; mean importance level for having a career was significantly (p. <.001) higher than the mean importance .evel for having a marriage and a family. Had the experiment included women who were not attending college, lhis effect may have achieved conventional levels of ignificance. 150 151 xxIIIIlIIIIIIIIIIIIIIIIIIIIIIIIII oo.ooH moH.moN mm Hoe.eeo.m~ Hmuoe eo.Hm moH.oo~ mm vom.m~m.mm uonum oo. HHmsm muw> ooo. H ooo. .o.s x .m.m x .m.m ow.H mom. Hom.H omm.mH¢ H cam-va .U.E N .m.m mo. mow. vow. one.ooH H mme.ooH .o.s x .m.m co. m. H0>O moo. MHh. H MHB. .mom N .m.m me. m. nm>o mmo. mmm.e H mmm.e monosumnoo rune Nw.N mOH. MNm.N 5mm.wa H hmmowmm OHDHHMM m0 Hawk mH.m mso. mmH.m moo.Hmo H omo.Hmo mHom xmm new.“ . awn. a new nmoe sueHHna uowumolleHucoouom cop MEHUmMIIm>oé autoimmune mm MHQMB xlIIIIIIIIIIIIIIIIIIIIIIIIIIHHHHull-IIIIIIIIIIIII- 152 Table 27 Estimated Career Ability--Means and Standard Deviations Fear of No Fear of Failure Failure Mean S.D. Mean S.D. Career 67.500 15.811 68.846 10.831 Congruent Condition Homemaking 59.000 11.370 60.000 15.275 Career 61.111 16.915 70.833 14.434 Incongruent Condition Homemaking 58.125 22.867 67.500 16.956 153 On the other hand, there was a significant (p < .009) sex role effect on the homemaking ability question (see Tables 28 and 29). The ability estimates of homemaking-oriented subjects were higher (X = 63.704) than those of career-oriented subjects (X = 52.791). This finding is consistent with the discovery that the importance of having a marriage and a family was signifi— cantly (p <.001) higher for subjects with a homemaking orientation than it was for subjects with a career orien- tation. While all subjects predicted that their per- formance would fall above the mean on the career test, career-oriented subjects with fear of failure predicted that their performance would fall slightly below the mean (X = 45.282) on a test of homemaking ability. This is an important finding which we shall expand upon shortly. When an analysis of the differences between predictions on the two tests was computed (see Tables 30 and 31), it was found that the differences, while in the expected direction, were significantly (p <.022) larger for the career-oriented subjects than for the homemaking— oriented subjects. Here again, we see the effects of our restricted sample which as a whole can be charac- terized as more career oriented than homemaking oriented. Up to now, the effect of fear of failure on these predictions has not been examined. For both the career test and the homemaking test the predictions of subjects 154 2// 00.00H mmh.hmv www.mmmrmv o o H“ mm mm moo mmm Nm th.ovm.mm #08 cm. m. Ho>o HNN. mmm.nm H mmm.nm .o HOHHW OH- x tho—m MW 0 I mo. m. Hw>o vmo. moH.hm H mVH.hm m w . .U.B x .m.m mm.m who 05H.m mHH.NmN.H H mHH.NmN~H .U..H. N .M.w hv.H mom. mHm.H OHN.mmm H OHN.mm m I v m m N .M.m HH. m Hw>o mHH. www.mv H www.mv wocwsumcou Emma hm.w «VNo. ovm.m www.mOH.N H mmN.mOH.N THSHHmm “O wam Nm.m emoo. me.h mmv.bmw~N H mmvohmw.m wHOM xwm mmuwswm . mumswm wouwsvm mo 85m w MHGmHm m cow: m0 m0 Edm mm andB umwa MHHHHnd wcHMMEOEomIImHHquOme cwquHummllm>ocd mmzlwonna 155 Table 29 Estimated Homemaking Ability--Means and Standard Deviations Fear of No Fear of Failure Failure Mean S.D. Mean S.D. Career 48.750 25.319 62.692 21.469 Congruent Condition Homemaking 55.667 13.211 63.077 17.974 Career 42.222 24.889 57.500 23.012 Incongruent Condition Homemaking 65.500 18.074 68.571 18.232 156 00.00H th.wom mm vmo.mmonom HMHOB Ho.mm ovo.mmm mm th.mmo.Hm HOHHm mH. m. H0>O hmH. Hhm.hm H Hum.hw .U.B x .m.m x .m.m hH.H vmm. mmN.H th.Non H th.Now .0.9 N .m.m ow. vmm. wmm. mNm.va H mNm.Hmv .O.B x .m.m mm. mom. who.H wvm.mmm H wvm.wam .m.m x .m.m mo. m. HO>O Hmo. wmm.mH H wmm.wH wocwsumcoo xmmB mm. mow. hon. www.mmm H www.mmm OHSHHMM mo Hmwm 0N.HH v...Hoo. mHH.NH mmm.vmhrm H mmm.¢mh.m wHom xmm H-HH . HHH.H. - ..-... huHHHnfi uwwumo mo puma so come so oHHucmouom cmumeHuwm om wHQwB can wuHHHnd mstcEmEom mo "whoom mocmHmMMHQIIm>os¢ mmzlmwuae 157 oo.ooH omo.mvm mm NHo.oom.mm Hmnoe MH.mm va.on mm Nom.ommrmm uouum mm.H «mm. on.H Hom.eHe H Hom.mHe .m.m x .m.m om. m. um>o Ham. HH~.NoH H HHN.~oH .o.e x .m.m om. Hee. mam. omm.omH H omm.omH .o.s x .m.m oo.m mmo. Neo.m mmo.Hmm.H H mmo.HNN.H .m.m x .m.m mv.m OHH. MHm.N mmb.omm H mmh.omm mocmsnucoo xmme mo. on. moo. mHo.mHm H mHo.NHN musHHmm Ho nmmm oH.m rmmo. mee.m oov.mme.H H ooe.mme.H mHom xmm .HHHH . HHH.H. .. HH.... wuHHHnd “condo mo umma so can muHHHnd mstwEwEom mo nonoom mocoHoMMHa mo mDHm> ousHomb¢lim>os¢ wmzlmenH ummy so mHHusmoumm cmumEHpmm Hm GHQMB 158 without fear of failure (career test X = 66.795, home- making test X = 62.960) were higher than those made by subjects with fear of failure (career test X = 61.434, homemaking test X = 53.535). This effect was significant only for the homemaking prediction (F = 5.340, p < .024); the F ratio did not achieve conventional levels of sig- nificance on the career test (F = 2.623, p < .109). These findings are consistent with the hypothesis sug- gested by the level of expectation data that subjects with fear of failure have less confidence in their abilities than subjects without fear of failure. A far more intriguing finding is the nearly significant (p < .053) sex role by fear of failure effect on the absolute value of differences between the two predictions. Career-oriented subjects with fear of failure tended to indicate more difference in ability level (X = 27.153) than did career-oriented subjects without fear of failure (X = 16.987). Homemaking-oriented subjects without fear of failure tended to indicate more difference in ability (x = 15.618) than did homemaking-oriented subjects with fear of failure (i = 11.438). Perhaps these results can resolve the contradic- tory findings reported by other experimenters on career salience in women. As noted by Almquist and Angrist (1970), explanations of female career salience have tended to employ either a deviance or an enrichment 159 hypothesis. Their data partially supported both hypothe- ses. Career salient subjects in their sample were sig— nificantly less likely to be going steady, engaged, or married as seniors or to belong to a sorority than sub— jects who were not career salient. On the other hand, career salient subjects were significantly more likely to have had working mothers than noncareer salient sub- jects. They also had had more and more various jobs and had been more influenced in their occupational choice by professors or members of the occupation than had non— career salient subjects. Kagan and Moss (1962) found significant associations between maternal hostility and acceleration of development during the first three years of life and the appearance of "masculine" interests in adulthood. They did note, however, that the raters used in their study may have misinterpreted the mothers' attitudes because of their subjective biases about normative child-rearing practices. It is interesting to note that Kagan and Moss view working mothers as nega- tive role models, while Almquist and Angrist view them as an enriching influence. White's (1959) and Lipman— Blumen's (1972) research tended to support the deviance hypothesis. The results of this study suggest that both hypotheses are valid. Some women (those without fear of failure) may choose a career orientation because they 160 are self-confident (significantly higher level of expec- tation, high predictions on both career and homemaking tests) and are encouraged by their successful parents (significantly higher socio—economic status) to actualize their goals directly rather than vicariously through their husbands. Other women (those with fear of failure) may choose a career orientation as part of a life-long pattern of high aspiration designed to avoid parental rejection. These high fear of failure women believe that they have less than average ability to perform tra— ditionally female role activities (X = 45.282). Perhaps these women come from families where the traditional female role is devalued and where their attempts to master female role activities were ridiculed. As a result they adopted a career orientation as a means of avoiding social rejection. While the difference between the two estimates was greater for the fear of failure subjects with a career orientation than for the career—oriented subjects without fear of failure, the opposite was true for the subjects with a homemaking orientation. The mean dif— ference for subjects with a homemaking orientation was greater for subjects without fear of failure than for subjects with fear of failure. Actually, the mean dif— ference for homemaking-oriented subjects without fear of failure (X = 15.618) was about the same as the mean 161 difference for career-oriented subjects without fear of failure (X = 16.987). The problem is to account for the small mean difference (2 = 11.438) of the fear of failure subjects with a homemaking orientation. It seems likely that this finding is a consequence of the generally low level of aspiration of fear of failure subjects with a hememaking orientation. They had the lowest mean esti- mate of ability for a test congruent with their sex role orientation; mean prediction was 61.6 for this group and 66.656 for the other three groups. In addition, the homemaking-oriented subjects with fear of failure had the lowest mean level of aspiration for the Scrambled Words Test. APPENDIX I ATTRIBUTION OF PERFORMANCE APPENDIX I Attribution of Performance Performance attribution is one of the few areas of achievement motivation research in which more than a few studies have included women. This research is based upon Heider's (1958) "naive analysis of action" which concludes that the disposition to perform an activity is a combination of the personal factors of effort and ability and of the environmental factors of luck and task difficulty. Feather (1968) has attempted to relate subjects' dispositional attributions of performance to their locus of control. More recently (1969), he has hypothesized that attribution is relatively more external for unexpected outcomes and relatively more internal for expected outcomes. This hypothesis has been confirmed in three experiments which included female subjects (Feather, 1969; Simon & Feather, 1973; Feather & Simon, 1974). It has also been reported (Feather, 1969; Simon & Feather, 1974) that female subjects are significantly more likely to attribute their performance to external (environmental) factors than are males. This finding is perplexing in light of Crandall, Katkovsky, and Crandall's (1965) discovery that girls in grades 6 through 12 claim significantly (p < .001) more personal responsi— bility for school grades than do boys. In another study 162 163 (Deaux & Emswiller, 1974) the focus was attribution of another person's performance. Once again, the con- nection between expectation and attribution was con- firmed; success on the masculine task was attributed to skill for males and luck for females (p. <.001). Males were also rated as significantly (p. <.05) more skillful than females. In the present study subjects were asked to indi- cate the percentage of their performance which they attributed to effort, skill, luck, and task difficulty. This question was included as an afterthought and was not an essential part of the experiment. Because the question was unclear and the significance levels of most of the effects discussed below only approach conventional levels of acceptability, these findings are to be regarded as highly tentative. Perhaps they can be best used to direct future research. The subjects were, unfortunately, not asked to indicate whether they felt they had done better or worse than they had expected, so analysis of the attribution responses is difficult. We shall proceed on the assumption that groups with mean post-performance estimates below the 50th percentile experienced their performance as a failure. Those groups whose mean post- performance percentile estimates were above the 50th Percentile must be regarded as being in an uncertain area. While they failed relative to their levels of 164 expectation, they still believed that their performance was above the average. Another problem affecting inter- pretation of these results is the fact that many subjects misinterpreted the instructions. A goodly number merely checked one of the four attributional factors, and two subjects gave attributions which added up to more than 100%. Given all these problems, perhaps the best method of examining these results is a chi square test. The category which received the largest numerical attribution was used as the classificatory category. Effort was seen as the most important factor by the majority of subjects (50), with skill second (17), task difficulty third (12), and luck fourth (4). Most of the subjects, therefore, took a large amount of personal responsibility for their performance. Internal factors were rated as the most important contributors significantly (X2 = 15.8, p. <.001) more often than the external factors. The direction of these effects were the same for the numerical data; mean attribution was greatest for effort (42.233), second for skill (25.908), third for task difficulty (19.901), and fourth for luck (11.844). Apparently the subjects cor— rectly perceived that performance (at least within this sample) was more a measure of effort than of ability. The four attribution factors were subjected to chi square analysis for the two subject variables. 165 While there were no significant differences in main source of attribution by sex role orientation, there was a significant (x2 = 8.17, df = 3, p. < .05) difference by fear of failure. Significantly (p < .05) more subjects without fear of failure considered skill the main source of their performance than did subjects with fear of failure. There was a tendency (p < .10) for more sub- jects with fear of failure to consider luck as the main source of their performance than did subjects without fear of failure. The numerical attribution scores were each separ- ately subjected to an unweighted means analysis of variance. The attribution to effort analysis (see Tables 32 and 33) produced no significant F ratios. There was, however, an insignificant (p < .077) fear of failure by task congruence interaction. This tendency becomes understandable when it is compared to the insignificant (p < .084) fear of failure by task congruence interaction for skill. Fear of failure subjects attributed to effort more in the congruent con- dition (congruent X = 52.271, incongruent X = 33.299) and to skill more in the incongruent condition (congruent X = 17.250; incongruent X = 25.052). In other words, they felt their effort (or lack of it) had a greater influence on performance in the congruent condition and that their ability (or lack of it) had a greater influence on performance in the incongruent condition. 166 ll|l||||lllllllll||||||| oo.ooH mmo.Hme.H we mNH.eso.moH Hence em.me mea.cme.H He mmc.moH.cm nouns Ho. m. Hm>o moo. oon.m H oo>.m .0.9 x .m.m x .m.m em.m see. men.m mam.smm.m H mmm.emm.m .o.s x .n.s -.N ceH. ocH.N amm.mm~.~ H amm.mmm.~ .o.s x .m.m oo. HHmEm huw> wHN. H wHN. .m.m N .m.m HH.H «on. Hoe.H moa.~eH.H H mea.meH.H moccshacoo race me. m. Hm:5 ewe. mmc.mm H mmc.m~ uHsHHcm no “can so. m. Hgo «so. Hmm.ae H Hmm.ee oHom rum umcmmmum .uHcon h dMMMWm so nwmcmwm uuomwm ou GOHuanHup¢I1m>os¢ hmzlomusa mm wHQMB 167 Subjects without fear of failure attributed their per- formance to effort more in the incongruent condition (X = 44.196) than in the congruent condition (X = Table 33 39.167). Attribution to Effort--Means and Standard Deviations Fear of No Fear of Failure Failure Mean S.D. Mean S.D. Career 46.875 26.584 33.333 30.625 Congruent Condition Homemaker 57.667 34.271 45.000 30.208 Career 37.222 30.116 48.750 40.121 Incongruent Condition Homemaker 29.375 29.770 39.643 34.276 It has previously been pointed out that fear of failure subjects had lower opinions of their abilities in general than subjects without fear of failure. Quite Possibly they expected to succeed on the congruent task and to fail on the incongruent task. If they perceived their performance as failing on both types of tasks (and this seems likely because their mean post-performance percentile estimate was 45.160), they experienced expeCted failure on the incongruent task and un failure on the congruent task. line with previous research, expected Their attributions, in suggest that they blamed 168 unexpected failure on the variable factor of effort and expected failure on the invariant factor of skill. Explaining the attributions of the subjects with- out fear of failure is somewhat more difficult because it is unclear whether this group experienced their per- formance as a success or as a failure. Since these subjects have a generally positive View of themselves, we might guess that any performance above the mean would be viewed by them as a success. AlthOugh they would expect to be successful on both tasks, their expectations would be higher for the congruent task. In line with previous research, less expected success (on the incon- gruent task) is more attributed to the variable factor of effort, while expected success (on the congruent task) is more attributed to the stable factor of ability. One significant F ratio was produced when the attribution to skill data was analyzed (see Tables 34 and 35), the sex role by task congruence interaction. This effect probably reflects the subjects' belief that a test of career ability demands more skill than a test of homemaking ability. For the career test, mean attri- bution to skill was 34.167 for the career—oriented sub- jects and 30.004 for the homemaking-oriented subjects. Attribution to skill for the homemaking test was lower-- 20.000 for the career-oriented subjeCtS and 19°462 for the subjects with a homemaking orientation. Another 169 oo.00H Nmm.mvh mm mmH.Hwotmh HMHOH mv.wm hNN.OHh Hm mmm.ommcvm HOHHH mv. m. Hw>o mmv. mmm.mNm H mmm.mNm .U.B x .m.m x .m.m mm.N vwo. Hmo.m wNm.MNH~N H mNm.mnH.N .U.B x .m.m Hm.v some. mmo.m omm.mmm~m H omm.mmm.m .0.9 x .m.m Ho. m. HO>O moo. omm.m H omm.m .m.m x .m.m HH. m. H0>O mOH. HHN.>> H HHN.hh mOCOSHOCOU Mch Hm.N hmo. mmm.N hhm.mNH.N H ohm.mNH.N OHHHHHcm mo Hmmm wH. m. HO>O mmH. Nmm.mNH H NNm.mNH OHOM xwm ..HHHH . ..HH. .. HH.... vm mHQcB HHH.H on 833837-865. music-cuss 170 interesting finding, which did not achieve conventional levels of significance (p.< .087), was that subjects without fear of failure attributed their performance to skill more (x = 30.665) than did subjects with fear of failure (X = 21.151). This finding is consistent with the interpretation that subjects with fear of failure view themselves as less skillful than do subjects without fear of failure. Table 35 Attribution to Skill--Means and Standard Deviations Fear of No Fear of Failure Failure Mean S.D. Mean S.D. Career 22.500 17.728 45.833 31.249 Congruent Condition Homemaking 12.000 15.213 26.923 29.898 Career 21.667 26.220 18.333 23.290 Incongruent Condition Homemaking 28.438 32.027 31.571 28.916 Analysis of the attribution to luck data (see Tables 36 and 37) produced two significant F ratios, the sex role by fear of failure interaction (p. <.008) and the three—way interaction (p <.026). Actually, both of these effects reflect the strong interaction of sex role and fear of failure in the incongruent condition. The attribution to luck made by career-oriented subjects 171 oo.o0H mVH.mHm mm omm.thsHm HMUOB Hm.om vhH.th Hm MHm.owo.mN HOHHH mm.v ammo. mNH.m mmm.OHv~H H mmm.OHv~H .U.H N .m.m N .m.m Hv.N NOH. mmn.~ omh.th H omh.mmh .U.B N .m.m hm.H NhH. mam.H vom.HNm H vom.HNm .U.B x .m.m mm.w smoo. va.b mmm.mvosm H mmm.mvorm .m.m x .m.m hN.H mmm. vvv.H mom.hmm H mom.hmm wocwduwcou meB mN.H 5mm. MNV.H vum.Hmm H vhm.Hmm OHSHHcm mo Hmmm mm.H hmH. HmN.N mmm.mHm H mmm.me wHom xww .HHHHH . HHH.H. .. H-.HH MOSH ou COHuanHuu¢I|c>os¢ wczlwmuaa mm OHQMB 172 with high fear of failure (X = 32.222) in this condition was nearly three times the size of the attribution to luck made by the next smaller group (X = 12.643) home— making-oriented subjects without fear of failure. The attributions to luck made by the other groups were nearly equal (X = 5.417 for career-oriented gs without fear of failure and X = 5.313 for homemaking—oriented subjects with fear of failure). Table 37 Attribution to Luck--Means and Standard Deviations Fear of No Fear of Failure Failure Mean S.D. Mean S.D. Career 10.000 11.650 10.000 11.481 Congruent Condition Homemaker 8.000 15.675 11.154 12.274 Career 32.222 39.220 5.417 9.876 Incongruent Condition Homemaker 5.313 10.242 12.643 13.703 To interpret these findings, we must examine the nearly significant (p < .056) three-way interaction for the attribution to task difficulty data (see Tables 38 and 39). In the incongruent condition, the attribution to task difficulty for fear of failure subjects with a homemaking orientation (X = 36.875) was much higher than that of fear of failure subjects with a career orientatlon oo. 00H Nme.ome om mem.mao.oe choa am.~m cHH.sae Ha mom.ecm.me nonnm mo.m who. mmn.m HHm.msm.N H HHm.mse.m .o.s x .m.m x .m.m so. m. nd>o mmo. oom.om H oom.om .o.s x .h.m mo. mme. Hmo. asa.mee H mem.~de .o.s x .m.m m oc. was. Ham. Hmm.dce H Hmm.mce .m.n x .m.m ms. m. nd>o ems. cme.oam H cme.oem docmsnmcoo rude am. mmv. Hum. eHd.eaa H sHm.ese dncHHdm no ncdm om.H How. oom.H mem.oHo.H H meN.cHo.H mHom xmm .HHHHH . HHH.H. .. HH.-H ansoHnnHo rues mm OHQMB on GOHuSQHnuuos¢ anglwonaa 174 (X = 10.000). If, as we have previously assumed, the fear of failure subjects experienced their performance on the incongruent task as expected failure, it appears that the SS with a homemaking orientation blamed their failure more on a fixed factor, task difficulty, while the S3 with a career orientation blamed their failure more on a variable factor, luck. This conclusion is consistent with our earlier discussion of the differences between career and homemaking-oriented subjects with fear of failure. The homemaking-oriented fear of failure subject believes she is generally of low ability, and her expected failure on the career task is a consequence of this lack of ability; i.e., the task was too difficult. The career-oriented subject with fear of failure, on the other hand, hopes that she is of generally high ability but has a lacuna in the homemaking area. Thus, although She expected to fail on this task, failure is generally inconsistent with her self-image. As a result, she blames her failure on the homemaking task on bad luck. Since we are unsure of the subjects' interpre- tations of their performance, we shall not examine the individual attribution SOurces further. As preViously mentioned, examination of the attributions along the internal—external dimension failed to produce any Sig— For all groups internal attribution nificant F ratios. was much greater (X = 67.966) than external attribution 175 (X = 31.764), suggesting that the subjects felt highly responsible for their performance. There was a non- significant trend for internal attribution to be greater for subjects without fear of failure (X = 72.347) than for subjects with fear of failure (if = 63.936) (p < .250) and for external attribution to be greater for subjects with fear of failure (X = 35.967) than for subjects without fear of failure (X = 27.523) (p < .203). These trends are suggestive of greater defensiveness on the part of subjects with fear of failure. As might be expected, there was a trend for external attribution to be greater in the incongruent condition (X = 35.724) (p.< .230) than in the congruent condition (X = 27.765) and for internal attribution to be greater in the congruent condition (X = 72.533) (p < .229) than in the incongruent condition (2 = 63.750). Table 39 Attribution to Task Difficulty--Means and Standard Deviations Fear of No Fear of Failure Failure Mean S.D. Mean S.D. Career 23.125 13.611 10.833 10.188 Congruent Condition Homemaker 36.875 41.708 17.929 19.117 Career 10.000 23.452 22.500 37.689 Incongruent Condition Homemaker 18.333 26.027 19.615 28.097 APPENDIX J HISTORY OF THE PROBLEM APPENDIX J History of the Problem It has frequently been asserted that the results of studies on achievement motivation in women are incon- sistent with each other and do not correspond to the findings in studies utilizing male subjects. This review of the literature will attempt to demonstrate that the reported inconsistencies are few and that the behavior of women in achievement situations is consistent with the expectancy-value theory of achievement motivation. The concept of motive to avoid success, it is hoped, will be revealed as unsubstantiated. Only a narrow range of achievement-related topics has been examined using female subjects. Much attention has been devoted to the arousal of achievement imagery and the performance of subjects differing in level of n Achievement, but almost no research has been conducted on fear of failure, level of aspiration, level of expec— tation, and preference for probability of success. Recently, considerable attention has been devoted to the motive to avoid success, and several studies have appeared which focus on an apparent decrement in per— formance which appears when women compete with men. 176 177 Arousal The earliest and most frequent concern of achievement motivation researchers working with female subjects was the arousal of n Achievement imagery. Field (1951) found that there was no difference in the amount of achievement imagery produced by his female subjects in a relaxed condition and in an achievement-aroused con- dition when the achievement arousal instructions empha- sized intellectual and leadership ability. When the T.A.T. n Achievement stimuli were administered after a highly arousing speech about social acceptability, how- ever, the level of n Social (his variable) was signifi— cantly greater than that produced in a relaxed condition. In contrast, his male subjects produced more n-Ach imagery after hearing the instructions which appealed to intel— lectual and leadership ability than they did in the relaxed condition, but their increase in n Social after social acceptability arousal did not achieve conventional levels of significance. These results suggest that in Field‘s University of Maryland sample the women measured their behavior against a standard of excellence which focused upon social behavior as the basis of competition, while the men compared their behavior chiefly against a standard of excellence which focused upon intelligence and leadership ability as the basis of competition. Note that both the male and the female subjects were 178 attempting to achieve a standard of excellence. Although some authors have suggested that the women in this study were attempting to gratify affiliation needs, a careful examination of the methodology indicates that social acceptability was described to the subjects as a skill essential to success in adulthood. The goal, success in adulthood, was the same for both sets of instructions. It appears that the men viewed intelligence and leadership ability as the skills necessary for their success in adulthood, while the women viewed social acceptability as the skill necessary for their success in adulthood. Field's study was conducted at a southern uni- versity during the height of the post-war "baby boom." Orso (1959), also working at a southern university (Susquehanna), found an increase in the amount of achievement imagery produced by his female subjects after affiliation arousal. Other studies have drawn their samples from populations which place greater or lesser emphasis upon social acceptability as a standard of excellence and/or view intelligence and leadership ability as necessary for female success in adulthood. A study by Veroff, Wilcox, and Atkinson (1953) revealed no significant differences in the n-Ach imagery produced by female subjects in a neutral or relaxed condition as compared to an achievement aroused condition which emphasized intelligence and leadership ability. Another 179 University of Michigan study (Fontana, 1970) found no significant differences between the amount of achievement imagery produced by female subjects in the neutral and the achievement aroused (intelligence and leadership ability) conditions. At the University of Michigan, as at the University of Maryland, women did not view intel- ligence and leadership ability as their standard of excellence. In Brazil (Angelini, 1955) where, during this time period, only extremely competitive women attended college, the n-Ach imagery produced by a group of coeds after administration of intelligence and leader— ship ability arousal instructions was significantly greater than that produced under neutral conditions. Lipinski, working with University of Cincinnati students in 1965, found an increase in the amount of achievement imagery produced by her subjects after intelligence and leadership ability arousal. Lesser, Krawitz, and Packard (1963) recognized that women may differ in the standard of excellence against which they measure their behavior. They found that gifted, female, high school achievers demonstrated a significant increase in the amount of n-Ach imagery produced to female stimuli after hearing intelligence and leadership ability arousal instructions. Under- achievers drawn from the same sample were found to have a significant increase in the amount of n-Ach produced 180 to male stimuli. The achievers produced less achievement imagery to the male stimuli after arousal than they had in the neutral condition, while the under-achievers pro- duced less n—Ach imagery to the female stimuli after achievement arousal. In a second (1964) study Lesser, along with Elizabeth French, specifically manipulated the arousal conditions and studied their interaction with the women's role orientation of the subjects' colleges. The dependent variable in this study was the number of achievement responses produced to French's Test of Insight. A highly significant interaction effect was discovered; women who attended colleges where the female students greatly valued intellectual achievement produced more achievement responses after intellectual arousal than after women's role arousal, but women who attended col— leges where the women greatly valued social and homemaking skill produced more achievement responses after women's role arousal than after intellectual arousal. Alper (1973) found that women with a nontraditional sex role ideology told significantly more stories in which the success of female figures was unambivalently valued than did women with a traditional sex role ideology. The stimuli in this study were administered under neutral conditions. Clearly, women, like men, demonstrate more achievement imagery after achievement arousal than in 181 a neutral or relaxed condition. The problem in comparing the results of the male studies to the results of the female studies has been that instructions emphasizing intelligence and leadership ability were assumed to be universally achievement arousing. In reality the standard of excellence valued by a group of individuals may vary within the group, across sexes, and within the same indi- vidual across time. It has consistently been found that women produce significantly more achievement imagery to male stimuli than they do to female stimuli (Field, 1951; Veroff, Wilcox, & Atkinson, 1953; Lesser, Krawitz, & Packard, 1963; French & Lesser, 1964; Lipinski, 1965; Cowan & Goldberg, 1967; and Wellens, 1973). This finding has usually been interpreted as indicating that women per- ceive achievement as more appropriate for men than for women. Even if this is the case, this finding in no way challenges the applicability of the expectancy—value theory to women. Perhaps men also perceive achievement as more appropriate for men than for women; this hypothesis has never, to the author's knowledge, been tested. Performance The results of studies using female subjects which relate n Achievement to performance have been consistent, despite popular belief to the contrary. 182 Veroff, Wilcox, and Atkinson (1953) found that in two middle time intervals the scrambled words performance of women with high n Achievement was significantly better than that of women with low n Achievement. Morrison (1954) also found that high n-Ach women performed sig- nificantly better on a scrambled words task than did women with low n-Ach. A significant relationship between n—Ach and skill attainment in swimming was found by Daugert (1966). More complicated results have been found when task description, type of subject, competition condition, type of feedback, and sex of competitor have been varied. It is quite clear that performance is enhanced when the task description emphasizes aspects of the task which are relevant to the subject's typical activities or goals. Milton (1959) found that women performed significantly better than men on arithmetic story problems which had typically female content. When the same arithmetic problems were written as stories with typically male content, the male subjects performed significantly better than the female subjects. French and Lesser (1964) attempted to demonstrate that female subjects perform better on a task described as relevant to their sex role orientation than on one irrelevant to their sex role orientation. Although the results confirmed their hypothesis, they must be discounted because two unrelated 183 performance measures were used in the two experimental conditions. Houts and Entwisle (1968) found that, when verbal ability was controlled, the combined English and Social Studies grades of 10th—grade girls who valued getting higher grades than boys were significantly higher than those of girls who did not value getting higher grades than boys. In another study (Maxwell and Gonzalez, 1972) subjects were asked to select endings for Vignettes which involved a woman in either a career or homemaking setting. For subjects with either a career or a homemaking orientation, there was a significant prefer— ence for the ending requiring less mastery on the task which was incongruent with the subject's sex role orien- tation. A number of studies have demonstrated that in situations which emphasize cooperation some women display a decrement in performance when they compete against a man whose prior performance seems to have been inferior to that of the woman. Weiss (1962) found that women dis— played a significant decrement in the amount of pressure exerted on a hand dynamometer when competing against a man relative to their performance when competing against a woman. It is important to note that the task in this experiment was described as an experiment about the acquaintanceship process. In another experiment in which subjects were asked to cooperate with each other (Morgan 184 & Mausner, 1973) female high school students who dis- covered that their embedded figure performance in a first session was superior to that of their male partners in the second session displayed a decrement in performance relative to their performance in the first session. The male subjects in this study who had had low scores in the first session displayed a relative increment in their performance. It should be noted that male subjects with high scores displayed a decrement in performance. These results suggest that in a cooperative situation women and, to a lesser extent men, try to equalize their per- formance with that of their partner. A study by Walker and Heyns (1962) suggests that women are more responsive than men to requests for decreased performance if the decreased performance will have the effect of increasing the success of the partner while having a negative effect on the subject's own task success. Fisher, O'Neal, and MacDonald (1974) found that success or failure feedback coupled with apparently unintentional feedback about the partner's liking or dislike for the subject affected per- formance on a second task. Female subjects with male partners displayed a relative decrement when they were accepted after failure or rejected after success. Sub— jects with female partners displayed a relative decrement after they were accepted after success or rejected after failure. The behavior of the subjects who had male 185 partners suggests that their goal was to be liked by the male partner, but the behavior of the subjects with female partners suggests a resistance to the desires of the partner. When the results of this study are combined with thOSe of Weiss's (1962) study, it appears that women in general reduce their performance to affiliate with males but affiliate through competition with females. The findings described above are consistent with the hypothesis that many women value following conven- tional patterns of female behavior more than they value task achievement. As previously suggested by the research of French and Lesser (1964) and Houts and Entwisle (1968), this tendency is apparent only in some women. Dickstein and Brown (1974) found that women with a traditional sex role orientation performed significantly lower on two W.A.I.S. subtests when they were told that their scores would be compared to those of men than they had on another subtest when told that they were being tested to develop norms for the test. Women with a nontraditional sex role orientation, on the other hand, performed better when they thought they were in competition with men than when they thought their performance would be used to develop norms. This finding is similar to Houts' and Entwisle's (1968) discovery that girls who value successful compe- tition with boys in school get better grades than girls who do not. 186 What conclusions can be drawn from these studies? It appears that women, when their attention is drawn to the cooperative aspects of a situation, tend to sacrifice their own performance for the sake of the joint effort. This pattern appears to be less characteristic of men than of women, but it is displayed by men of superior ability (Crandall, Katkovsky, & Preston, 1969; Morgan & Mausner, 1973). In an ambiguous situation some women (those with a nontraditional sex role orientation) seem to opt for a superior position relative to a male com- petitor, while others (those with a traditional sex role orientation) seem to opt for an inferior position relative to a male competitor. Perhaps women in the first group feel that they should try their best on every task, while ‘ the women in the second group feel that women should appear to be inferior to a male competitor when competing in a traditionally male realm. These stances suggest different standards of excellence. Motive to Avoid Success Horner (1968, 1971, 1972, 1973a, 1973b) has hypothesized that some women have a motive to avoid success which inhibits their performance in competition. She postulated "a woman is threatened by success because unusual excellence in academic intellectual areas is unconsciously equated with loss of femininity, the con- sequence of which may be social rejection (1968, p. 16)." 187 This feared loss of femininity, she suggested, is a con— sequence of the aggressive overtones of success via com- petition and is thus more likely to be active when com— petition is against another individual than when the competition is against an internal standard. She developed a scoring system for motive to avoid success and applied it to stories told to the stimulus, "After first term finals, Ann finds herself at the top of her medical school class.“ The major findings of her study relative to motive to avoid success were that women with fear of success performed better when working alone than when working in competition and that women demon— strated more fear of success than men. The first result seems invalidated by Burghardt's (1973) finding that women with fear of success performed significantly better on an anagram task, whether working alone or in compe— tition, than did women without fear of success. Feather and Simon (1973) found no significant differences in performance in a group setting between women who told fear of success stories and those who did not. This finding also does not support Horner's contention that fear of success inhibits performance in competition. Several studies have found the same or a higher incidence of fear of success imagery among men than women when the "Anne" stimulus was used (Morgan & Mausner, 1973; Burghardt, 1973; Feather & Raphelson, 1974; Monahan, 188 Kuhn, & Shaver, 1974). The incidence of fear of success imagery has been found to vary widely with the sample, sex of stimulus figures, and content of the stimulus (Weston & Mednick, 1970; Kresjevich, 1972; Burghardt, 1973; Breedlove & Cicirelli, 1974; Feather & Raphelson, 1974; Jackaway, 1974; Tomlinson-Keasey, 1974). In spite of these indications that the concept is invalid or at least that the technique for assessing the motive is unreliable, considerable research has utilized fear of success as an independent variable. Moore (1972) found no relationship between fear of success and female curricular choice in graduate school. Hertzog and Walker (1973) failed to discover any relationship between experi- menter bias and fear of success. The incidence of fear of success stories was found to be greater in pre-menstrual than inter-menstrual women by Patty and Ferrell (1974). Good and Good (1973) have developed a questionnaire measure of motive to avoid success but have not, up to this point, demonstrated its validity. There have been some interesting research find- ings relative to fear of success which are probably to some extent a function of the atypical (for a woman) success depicted in the "Anne" stimulus. Makosky (1972) found that women with fear of success performed sig- nificantly better in competition with a woman or when working alone than they did when competing against a man. 189 Women without fear of success performed significantly better when competing against a man than they did when working alone or competing against a woman. A study by Karabenick and Marshall (1974) failed to replicate this relationship. Makosky (1972), Tomlinson-Keasey (1974), and Feather and Raphelson (1974) suggest that fear of success responses reflect the subject's feeling that Anne's success is inconsistent with her conception of appropriate female role behavior. Parker (1971) found that women with high fear of success were significantly more concerned with homemaking than women with low fear of success. Women with low fear of success were significantly more inter- ested in having a career than were women with high fear of failure. On a scrambled words task, high fear of success subjects performed significantly better when the task was described as feminine than when it was described as masculine, while low fear of success subjects per- formed significantly better On the task when it was described as masculine than when it was described as feminine. It appears that women who tell fear of success stories to the "Anne" stimulus favor competition against women (Makosky, 1972) on traditionally feminine tasks, while women who do not tell fear of success stories to the "Anne“ stimulus favor competition against men on traditionally male activities. Horner, herself, reports 190 (1972) that 88.9% of the fear of success subjects in her original (1968) study were majoring in the humanities (a traditional refuge of noncareer-oriented women), while 56% of the low fear of success women were majoring in the natural sciences. There appears to be little evidence for Horner's assertion that fear of success is a motive. Fear of Failure As pointed out by Stein and Bailey (1973) and Karabenick and Marshall (1974), little attention has been devoted to the effect of motivation to avoid failure on achievement behavior in women. The belief that the expectancy—value theory is not applicable to women is probably at least partially responsible for this defi- ciency. Then, too, there have been several studies utilizing the test anxiety measure of fear of failure which have failed to produce significant results. Daugert (1966) found no significant relationship between test anxiety and acquisition of swimming skill. Horner (1968) found no significant relationship between Debilitating Achievement Anxiety scores and performance for her female subjects. A Norwegian study (Gjesme, 1973) of seventh— grade girls revealed no significant relationship between the school grades of girls with the four different combinations of n-Ach and test anxiety. Issacson (1964) was unable to find significant differences in difficulty of area of curricular choice between women with high n—Ach 191 and low test anxiety and women with low n-Ach and high test anxiety. On the other hand, King—Fun Li (1974), working in Hong Kong, found that high levels of test anxiety were associated with certain parental attitudes. The author was able to find only four studies with female subjects which used a projective measure of the motive to avoid failure. Vollmer (1973), working in Norway, found a significant negative correlation between a Holtzman ink blot measure of personality definition and fear of failure as measured by Heckhausen's system. Three validation studies for the hostile press scoring system for fear of failure add to the impression that fear of failure in women has a negative relationship to personality integration. Hostile press imagery was found to have a significant negative relationship to self-ideal congruence as measured by a Q-sort (Smith & Teevan, 1964). When related to the F scale, hostile press imagery was found to have a significant negative relationship with anti-intraception and a significant positive relation- ship to destructiveness and cynicism (Teevan & Hartsough, 1964). In the same study it was found that hostile press imagery had a significant positive relationship to the aesthenic sub-scale of the S~V scale. In a related study (Teevan & Hartsough, 1964) hostile press imagery was found to have a nearly significant negative correlation with the MF scale of the MMPI. In other words, women 192 with high levels of hostile press imagery tend to have traditionally female interests and attitudes. Also in the same study it was found that, as measured by the A-S reaction study, high hostile press imagery women tended to be ascentive. Hostile press imagery was also found to have a significant association with manifest anxiety. Because Teevan found that hostile press imagery was associated with a different pattern of personality variables in women than in men, no further research using this measure was conducted. The importance of investigating the effects of fear of failure on achievement behavior in women has recently been reinforced by the findings of Karabenick and Marshall (1974). Using the Haber-Alpert Achievement Anxiety Test, these experimenters found a significant three-way interaction between fear of failure, fear of success, and type of opponent, on improvement in a digit substitution task. Subjects with low fear of success and low fear of failure and those with high fear of success i and high fear of failure improved most when competing against a man. Subjects with low fear of success and high fear of failure improved most when competing against a woman, while those with high fear of success and low fear of failure improved most when working alone. The meaning of these findings is unclear for two reasons: (1) As previously discussed, fear of success seems to 193 reflect sex role stereotypes rather than motivation level, and (2) The Achievement Anxiety Test scores may reflect anxiety about the possible negative consequences of success as well as the possible negative consequences of failure. A more meaningful finding in this study was that subjects with high fear of failure improved to a signifi— cantly greater extent after success feedback than did subjects with low fear of failure. Subjects with low fear of failure improved significantly more after failure feedback than did subjects high in fear of failure. These findings are important because they are similar to the findings of studies using male subjects. Addi- tional findings of interest were that high fear of failure subjects were significantly more affected by feedback and thought others were significantly more affected by their performance than were low fear of failure subjects. Clearly these findings are only a beginning. REFERENCES REFERENCES Almquist, E. M., & Angrist, S. S. Career salience and atypicality of occupational choice among college women. Journal of Marriage and the Family, 1970, 32, 242-249. Alper, T. G. The relationship between role orientation and achievement motivation in college women. Journal of Personality, 1973, 41, 9-31. Alpert, R., & Haber, R. N. Anxiety in academic achieve- ment situations. Journal of Abnormal and Social Psychology, 1960, 61, 207-215. Angelini, A. L. Um novo metodo para avaliar a motivacao humano, Bol. Fac. Filos. Cienc. S. Paulo, 1955, No. 207. Cited by Lesser, G., Krawitz, R. N., & Packard, T., Experimental arousal of achievement motivation in adolescent girls, Journal of Abnormal and Social Psychology, 1963, 66 (1), 59-66. Atkinson, J. W. (Ed.) Motives in fantasy, action, and society. Princeton, N.J.: D. Van Nostrand, 1958. Atkinson, J. W. Motivational determinants of risk-taking behavior. In J. W. Atkinson and N. T. Feather (Eds.), A theory of achievement motivation. New York: JOhn Wiley & Sons, 1966. Atkinson, J. W., & Feather, N. T. A theory of achievement motivation. New York: John Wiley & Sons, 1966. Atkinson, J. W., & O'Connor, P. Neglected factors in studies of achievement-oriented performance. In J. W. Atkinson and N. T. Feather (Eds.), A theory of achievement motivation. New York: John Wiley & Sons,71966, 299-325. Bardwick, J. M. Psychology of women. New York: Harper & Row, Puinshers, 1971. 194 195 Birney, R. C., Burdick, H., & Teevan, R. C. Fear of failure. New York: Van Nostrand—ReinhoId Company, 1969. Breedlove, C. J., & Cicirelli, V. G. Women's fear of success in relation to personal characteristics and type of occupation. The Journal of Psy— chology, 1974, 86, 181-190. Broverman, I. K., Vogel, S. R., Broverman, D. M., Clark- son, F. E., & Rosenkrantz, P. S. Sex role stereo- types: A current appraisal. The Journal of Social Issues, 1972, 28, 59—78. Burghardt, N. R. The motive to avoid success in elemen- tary school students. Paper presented at the meeting of the American Psychological Associ- ation, Montreal, Canada, August, 1973. Cass, J., & Birnbaum, M. Comparative guide to American colleges. New York: Harper & Row Publishers, Inc., 1972. College Entrance Examination Board. Your college board scores: Scholastic aptitude test, achievement test. Princeton, N.J.: College Entrance Exami— nation Board, 1963. Cowan, G., & Goldberg, F. J. Need achievement as a function of the race and sex of figures of selected TAT cards. Journal of Personality and Social Psychology, 1967, 5, 245—249. Crandall, V. C. Sex differences in expectancy of intel- lectual and academic reinforcement. In C. P. Smith (Ed.), Achievement-related motives in children. New York: RusseII Sage, 1969. Crandall, V. C., Katkovsky, W., & Crandall, V. J. Children's beliefs in their own controls of reinforcements in intellectual-academic achieve- ment situations. Child Development, 1965, 36, 91-109. Daugert, P. J. The relationship of anxiety and the need for achievement to the learning of swimming. Unpublished doctoral dissertation, University of Michigan, 1966. 196 Deaux, K., & Emswiller, T. Explanations of successful performance on sex—linked tasks: What is skill for the male is luck for the female. Journal of Personality and Sooial Psychology, 1974, 62, 80185. DeCharms, R. Personal causation. New York: Academic Press, 1968. DeCharms, R., & Dave', P. N. Hope of success, fear of failure, subjective probability, and risk-taking behavior. Journal of Personaligy and Social Psy— chology, 1965, A, 558-568. Dickstein, L. S., & Brown, N. Effect of role orientation and instructions regarding competition on cog- nitive performance of college females. Ps - chological Reports, 1974, 66, 291-297. Diggory, J. C., & Morlock, H. C., Jr. Level of aspiration or probability of success? Journal of Abnormal and Social Psychology, 1964, 62, 282-289. Entwisle, D. R. To dispel fantasies about fantasy-based measures of achievement motivation. Psychological Feather, N. T. Performance at a difficult task in relation to initial expectancies of success, test anxiety, and need achievement. Journal of Personality, 1965a, 66, 200-217. Feather, N. T. Valence of success and failure in relation to task difficulty: Past research and recent progress. Australian Journal of Psychology, 1968, 66, 111-122. Feather, N. T. Attribution of responsibility and valence i of success and failure in relation to initial confidence and task performance. Journal of Personality and Social Psychology, 1969, I; (2), 129-144. Feather, N. T., & Raphelson, A. C. Fear of success in Australian and American student groups: Motive or sex—role stereotype? Journal of Personality, 1974, 63, 190-201. Feather, N. T., & Simon, J. G. Fear of success and causal attribution for outcome. Journal of Personality, 1973, 61, 525-542. 197 Field, W. P. The effects of thematic apperception of certain experimentally aroused needs. Unpub- lished doctoral dissertation. University of Maryland, 1951. Fisher, J. E., O'Neal, E. C., & McDonald, P. J. Female competitiveness as a function of prior performance outcome, competitors evaluation, and sex of com- petitors. Paper presented at the meeting of the Midwestern Psychological Association, Chicago, May 1974. French, E. G., & Lesser, G. S. Some characteristics of the achievement motive in women. Journal of Abnormal and Social Psychology, 1964, 66(2), 119-128. Gjesme, T. Achievement-related motives and school per- formance for girls. Journal of Personality and Social Psychology, g6_(1), 131-136. Good, L. R., & Good, K. C. An objective measure of the motive to avoid success. Psychological Reports, 1973, 66, 1009-1010. Hancock, J. G., & Teevan, R. C. Fear of failure and risk- taking behavior. Journal of Personality, 1964, 66, 200-209. ‘ Heckhausen, H. The anatomygf achievement motivation. New York: Academic Press, 1967. Heider, F. The psychology of interperggnal relations. New York: John Wiley & Sons, Inc., 1958. Hertzog, J., & Walker, C. E. Effects of sex and need to avoid success on verbal mediation of experimenter bias. Psychological Reports, 1973, 62, 1235-1238. Horner, M. S. Sex differences in achievement motivation and performance in competitive and non-competitive situations. Ann Afbor: University Microfilms, Inc., 1968. Horner, M. S. Femininity and successful achievement: A basic inconsistency. In J. M. Bardwick, E. Douvan, M. S. Horner, and D. Gutmann (Eds.), Feminine personality and conflict. Belmont, CaIif.:—Wadsworth‘Publis51hg Company, Inc., 1970, 45-76. 198 Horner, M. S. The psychological significance of success in competitive achievement situations. A threat as well as a promise. In H. I. Day, D. E. Ber— lyne, and D. E. Hunt (Eds.), Intrinsic motivation: A new direction in education. Toronto, Canada: Holt, Rinehart and’Winston of Canada, Limited, 1971, 46-60. Horner, M. S. The motive to avoid success and changing aspirations of college women. In J. M. Bardwick (Ed.), Readings on the psychology of women. New York: Harper & Row, Publishers, 1972a, 62-67. Horner, M. S. Toward an understanding of achievement- related conflicts in women. Journal of Social Horner, M. S. Scoring manual for an empirically derived scoring system for motive to avoid success. Paper presented at the meeting of the American Psycho— logical Association, Montreal, Canada, August, 1973. Horner, M. S., & Walsh, M. R. Causes and consequences of existing psychological barriers to self-actuali- zation. Annals of the New York Academy of Sciences, 1973, 666, 124-130. Houts, P. S., & Entwisle, D. R. Academic achievement effort among females: Achievement attitudes and sex—role orientation. Journal of Counseling Psychology, 1968, $6, 284—286. Issacson, R. L. Relation between n achievement, test anxiety, and curricular choices. Journal of Abnormal and Social Psychology, 1964, 66, 447-452. Jackaway, R. Sex differences in the development of fear of success. Child Study Journal, 1974, 6, 71—79. Kagan, J., & Moss, H. Birth to maturity. New York: John Wiley & Sons, Inc., 1962. Karabenick, S. A., & Marshall, J. M. Performance of females as a function of fear of success, fear of failure, type of opponent, and performance contingent feedback. Journal of Personaligy, 1974, 62, 220-237. King—Fun Li, A. Parental attitudes, test anxiety, and achievement motivation: A Hong—Kong study. Journal of Social Psychology, 1974, 66, 3-11. 199 Klinger, E. Fantasy need achievement as a motivational construct. Psychological Bulletin, 1966, 66 (4), 291-308. Klinger, E. Short-term stability and concurrent validity of TAT need scores: Achievement, affiliation, and hostile press. Proceedings of the 76th Annual Convention of the American Psychological Association, 1968, 157-158. Klinger, E., & McNelly, F. W., Jr. Fantasy need achieve— ment and performance: A role analysis. Psy— chological Review, 1969, 16, 574-591. "Kresjevich, I. 2. Motivation to avoid success in women as related to year in school, academic achievement, and success context. Unpublished doctoral disser— tation, Michigan State University, 1972. Kukla, A. Foundations of an attributional theory of per- formance. Psychological Review, 1972, 16, 454- Lesser, G., Krawitz, R. N., and Packard, R. Experimental arousal of achievement motivation in adolescent girls. Journal of Abnormal and Social Psychology, 1963, 66 (1), 59-66. Lipinski, B. G. Sex role conflict and achievement moti- vation in college women. Dissertation Abstracts, 1966, E6, 4077. Lipman-Blumen, J. How ideology shapes women's lives. Scientific American, 1972, 226 (1), 34-42. McClelland, D. C., Atkinson, J. W., Clark, R. A., & Lowell, E. L. The achievement motive. New York: Appleton-Century-Créfts, 1953. Makosky, V. P. Fear of success, sex—role orientation of the task, and competitive condition as variables affecting women's performance in achievement— oriented situations. Paper presented at the meeting of the Midwestern Psychological Associ— ation, Cleveland, Ohio, 1972. Cited by S. A. Karabenick and J. M. Marshall, Performance of females as a function of fear of success, fear of failure, type of opponent, and performance feedback. Journal of Personality, 1974, 42, 220—237. - 200 Mandler, G., & Sarason, S. B. A study of anxiety and learning. Journal of Abnormal and Social Psy— chology, 1952, 61, 166-173. Maxwell, P. G., & Gonzalez, A. E. J. Traditional and non- traditional role choice and need for failure among college women. Psychological Reports, 1972, 66, 545—546. Mettee, D. R. Rejection of unexpected success as a function of the negative consequences of accept- ing success. Journal of Personality and Social Psychology, 1971, 61, 332-341. Milton, G. A. Sex differences in problem solving as a function of role appropriateness of the problem content. Psychological Rgports, 1959, 6, 705-708. Monahan, L., Kuhn, D., & Shaver, P. Intrapsychic versus cultural explanations of the "fear of success" motive. Journal of Personaligy and Social Psy— chology, 1974, 66, 60-64. Moore, L. L. The relationship of academic group member- ship to the motive to avoid success in women. Dissertation Abstracts, 1972 (Feb.), 32 (8-A), 4355. Morgan, S. W., & Mausner, B. Behavioral and fantasied indicators of avoidance of success in men and women. Journal of Personality, 1973, 66, 457‘470. Morrison, H. W. Validity and behavioral correlates of female need for achievement. Unpublished Master's thesis, Wesleyan University, 1954. Cited by D. C. McClelland (Ed.), Studies in motivation. New York: Appleton-Century—Crofts, Inc., 1955. Orso, D. P. Comparison of achievement and affiliation arousal on n-Ach. Journal of Projective Tech- niques and Personality Assessment, 1969, 66 (3), 230—233. Parker, V. J. Fear of success, sex-role orientation of the task, and competition condition as variables affecting women's performance in achievement- oriented situations. University Microfilms, 1971. Patty, R. A., & Ferrell, M. M. A preliminary note on the motive to avoid success and the menstrual cycle. Journal of Psychology, 1974, 66, 173-177. 201 Raynor, J. 0. Future orientation and motivation of immediate activity: An elaboration of the theory of achievement motivation. Psychological Review, 1969, 16, 606-610. Reiss, A. J., Jr., Duncan, 0. D., Hatt, P. K., & North, C. C. Occupations and social status. New York: The Free Press, 1961. Ryan, E. D., & Lakie, W. L. Competitive and noncompetitive performance in relation to achievement motive and manifest anxiety. Journal of Personality and Social Psychology, 1965,4342-345. Sarason, I. G. The effects of anxiety and threat on the solution of a difficult task. Journal of Abnormal Psychology, 1961, 66, 165-168. Schrauger, J. S., & Rosenberg, S. E. Self-esteem and the effects of success and failure feedback on per- formance. Journal of Personality, 1970, 66, 404-417. Schwenn, M. Arousal of the motive to avoid success. Unpublished junior honors thesis, Harvard Uni- versity, 1970. Cited by M. S. Horner, Toward an understanding of achievement—related conflicts in women. Journal of Social Issues, 1972, 66 (2), 157-175. Simon, J. F., & Feather, N. T. (Causal attributions for success and failure at university examinations. ggurnal of Educational Psychology, 1973, 66, 46-56. Smith, B. D., & Teevan, R. C. Relationships among self- ideal congruence, adjustment, and fear of failure motivation. gpurnal of Personalipy, 1971, 66 (1), 44-56. Stein, A. S., & Bailey, M. M. The socialization of achievement orientation in females. Psycho- logical Bulletin, 1973, fig. 345-366. Teevan, R. C., & Hartsough, W. R. Personality correlates of fear of failure vs. need achievement individual. I. value scales. II. a clinical picture. Unpub- lished Office of Naval Research Technical Report No. 5, Contract Nonr 3591(01), NR 171-803, 1964. 202 Teevan, R. C., & McGhee, P. E. Childhood development of fear of failure motivation. Journal of Per- sonality and Social Psychology, 1 , _6 3), 345—348} Terman, L. M., & Oden, Volume 5: ford? M. H. Genetic studies of genius, The gifted group at mid-life. Stan- Stanford University Press, 1959. Teghtsoonian, M. Distribution by sex of authors and editors of psychological journals, 1970-1972 "Are there enough women editors?" American Psychologist, 1974, 66, 262—269. Tomlinson—Keasey, C. Role variables: Their influence on female motivational constructs. Journal of Counseling Psychology, 1974, 66, 232-237. Thomas, C., & Teevan, H. C. Level of aspiration and motive patterns, Technical Report No. 2, 1964. Cited by R. C. Birney, H. Burdick, and B. C. Teevan, Fear of failure. New York: Van Nostrand-Reinhold Company, 1969. 0.8. Bureau of the Census. Statistical abstract of the United States: 1973. Washington, D.C., 1973. Veroff, J., Wilcox, S., & Atkinson, J. W. The achieve- ment motive in high school and college age women. The Journal of Abnormal and Social Psychology, 1953, 66 (1), 108—119. Veroff, J. Social comparison and the development of achievement motivation. In C. P. Smith (Ed.), Achievement—related motives in children. New Yofk: RusseII Sage Foundation, 1969, 46-101. Vollmer, F. Achievement motivation and perception of the body boundary. Scandinavian Journal of Psy- chology, 1973, 66, 91-95. Walker, E. L., & Heyns, R. W. An anatomy for conformity. Englewood, N.J.: Prentice-Hall, Inc., 1962. Weiner, B. The effects of unsatisfied achievement moti- vation on persistence and subsequent performance. Journal of Personality, 1965, 66, 428-442. Weiss, P. Some aspects of femininity. Dissertation Abstracts, 1962, 66, 1083. 203 Wellens, G. The motive to avoid success in high school seniors: N—achievement shifts and psychosocial correlates. Dissertation Abstracts International, 1973 (May), 33 (ll-B), 5529. Weston, P. J., & Mednick, M. T. Race, social class and the motive to avoid success in women. Journal of Cross-Cultural Psychology, 1970, 6 (3), 2 White, B. J. The relationship of self concept and parental identification to women's vocational interests. Journal of Counseling Psychology, 1959, 6, 202- 206. Ziskin, M. L. Relationship of achievement motivation to perception of degree of task difficulty and estimate of performance. Dissertation Abstracts, 1966, 66_(12, pt. 1), 7453. Supplementary References Alper, G. Predicting the direction of selective recall: Its relation to ego strength and n achievement. Journal of Abnormal and Social ngchology, 1957, 66, 149-165. Alper, G. Achievement motivation in college women: A now-you—see-it-now-you-don't phenomenon. American Psychologist, 1974, 66, 194-203. Alper, G., & Greenberger, E. Relationship of picture structure to achievement motivation in college women. Journal of Personality and Social Psy- chology, I967, 67’362-37I. Atkinson, J. W. Studies in projective measurement of achievement motivation, Microfilm Abstracts, Douvan, E., & Adelson, J. The adolescent experience. New York: John Wiley & Sons, Inc., 1966. Feather, N. T. The relationship of expectation of success to n achievement and test anxiety. Journal of Personality and Social Psycholggy, 19 , _, 118-126. Frank, J. D. Level of aspiration test. In H. A. Murray (Ed.), Explorations in pers onalit . New York: Oxford University Press, 1938, 461—471. 204 Frankel, P. M. Littig, Lunneborg, P. W., McCoy, N. Mead, M. Smith, B. D., Stewart, A. J., Sex-role attitudes and the development of achievement need in women. Journal of College Student Personnel, 1974, 66, 114-119. L. We] & Yercaris, C. A. Academic achievement correlates of achievement and affiliation moti— vations. The Journal of Psychology, 1963, 55, 115-119 . ‘— & Rosenwood, L. M. and achievement: Need affiliation Declining sex differences. Psycholggical Reports, 1972, 66, 795—798. Effects of test anxiety on children's per- formance as a function of instructions and type of task. Journal of Personality and Social Psychology, 1965, 6, 634-641. Male and female. New York: Dell Publishing Co., Inc., 1919. & Teevan, R. C. The relationship of the hostile press measure of fear of failure to self- ideal congruence and adjustment. Unpublished Office of Naval Research Technical Report No. 11, Contract Nonr 3591(01), NR 171-803, 1964. & Winter, D. G. Self definition and social definition in women. Journal of Per- sonality, 1974, 66, 238—259. Springfield? Mass.: Webster‘s New Collegiate Dictionapy, Second Edition. G. & C. Merriam Co., Publishers, 1959. ()(ijmu) l (um (j fill/Ml S E I R AH R B I L V T. N U E T AH TI 3 N ha G T. H C I w.