' w-v r-Pr ::".‘II *. ‘1 II!‘.? I .I'I" . #3:" . '| g. I I ' I~ I': | 3. v fl I1§FIIII 11¢ : J1} I‘lé‘ *5 r' :1; ”I‘. “I11."'1Ilf;l.;: I 'I.‘I‘.".‘ (III ,g‘II‘H 5 I «If»; <45 . 3 1" :I‘ I1: '1"- I II 353 ' ' .qu 1,}. ~.‘ 2‘“ I I vIr' 151M». ' I “.1I'.'I"I'IIIII‘.I3I’,.-I 13:55?” A. I ‘ . II. 1 " . . I - . ; ‘ J 1..» - A 7 ' '31.:— 1:11,” ._ I-.. {III- IAIIII' I, 311 . I I ‘) “A“ H5“), “31:13?" ““611 _.,.;_‘;5'1-‘1.1hr.~lll'IL'L‘Z'IM .I .h ' .6“ If) (-5-, 15-??1’fi . 1,4'NV1I 4 3$%°?- 4b “fir “' _1 a' \‘*a3h ‘T3225 .;;;3xI':C' 8C3. I #3527}, ‘II'rl v"\>g')‘ H‘AIII I‘l‘l* 'ailI :31 {an}? JIOE: , I . ‘;Ia‘ .‘II.< I I .‘ w ‘ . '-. " ‘ -I._ '. n .1.‘}LII1L..IZ' ' .,,»'|.'11:., ‘ . .2, I, f“ - 5 g g; 5"”; ‘ f‘fi'l§.¥-§7‘:§«%{W «>- . ,‘Ju-L ‘1"l‘l‘lt.r" '1 . ~_I , Iv HI.) ‘I 1|] ‘ ‘ ‘ I 4 .- wg :v;.wy$—53I',I.-n:-..‘ .39.“? . -_€_".‘.:\—':._ ‘:'III if“. .5‘ W,.IE_,3‘*;A::\;I __'__ I :57; I Jr :IXI- -‘I“F‘i ”1" .". . ' I I :75. ('11.:- 5P“ 3'12; 5‘ 5, u I'. 'dIu'A‘ . ;; . : '5 3. ' 5I .51! . - ‘ . II I 5; I» ‘ ~ - I‘ "" {In 1I 1|;I1.11¢11I131II 1-36.}! I I I I. IIII'WQSI 1"; ‘fmkf .131“nkffi 1' i' ' ,... .fl_l _ 1 .3 ‘5‘1 U .I ‘ ' . 5.]. ' I I ¢| 5 I5 I' I " u ‘1 II ‘ I: ”1'... 1 ‘ II ' ," ‘ 11'. ' 1 . "0"1} "1-5 IIIIIIM IIl'IIIV'II'1M""I‘ I". '11 11551 M ‘y(.u;1’151rvIII‘I'Id5 'I1I1II1 II, ‘ 11,“; Date 0-7639 l mu» l m mm l‘l'l [lull Illlllljjlfll 293 Kim r auMe~o.‘.‘...-.¢.- g I. III 3? A R. Y M l ham State E - i - L} L1 my ersity f . f o'er f 'flr «‘n ”3"" "9 “i‘omfi This is to certify that the thesis entitled The MWPI MF Scale in College Women -- An Empirical Investigation of Some Clinical Assumptions presented by Nancy Jeanne Egan has been accepted towards fulfillment of the requirements for Ph.D. degree in Clinical Psychology 41/ ZC¢404444 . g /, Major professor / .3—//,; —X’/ OVERDUE FINES: 25¢ per day per item RETURNIM LIBRARY MTERIALS: Place in book return to Miove charge from circulation records JEWEL" THE MMPI MF SCALE IN COLLEGE WOMEN-- AN EMPIRICAL INVESTIGATION OF SOME CLINICAL ASSUMPTIONS BY Nancy Jeanne Egan A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1980 ABSTRACT THE MMPI MF SCALE IN COLLEGE WOMEN-- AN EMPIRICAL INVESTIGATION OF SOME CLINICAL ASSUMPTIONS BY Nancy Jeanne Egan The Masculinity-Femininity Scale of the MMPI consists of 60 items related to interests, occupational choices, emotional sensitivities, and aesthetic preferences. This dissertation examined the assumption that low T scores for women on the Mf Scale are associated with submissive and passive trends in the personality. Another assumption under consideration was whether or not high T scores on Scales l (Hypochondriasis), 2 (Depression), and 3 (Hysteria) are associated with low Mf Scale T scores. Specifically, the present study hypothesized that low Mf Scale scores would be associated with low scores on a measure of interpersonal assertiveness. It was further hy- pothesized that if the Mf Scale was a valid measure of femi- ninity subjects with low T scores would be feminine on another measure of sex role orientation. To test these hypotheses 419 college women were ad- ministered the MMPI, the College Self-Expression (Galassi, DeLo, Galassi, & Bastien, 1974) as a measure of assertive- ness, and the Bem Sex Role Inventory (Bem, 1974). Scores on the Mf Scale were categorized into high and low groups Nancy Jeanne Egan using two research methodologies: (1) a quartile method considering the Highest ("masculine") and Lowest (feminine") Quartiles on the Mf Scale based on the Mf scores of the entire sample (Hathaway & Meehl, 1950); and (2) a Highest Point-Lowest Point method based on the individual's highest and lowest T scores of the profile, i.e., a "masculine" Mf group consisting of those with the peak T score on Mf, and a "feminine" Mf group of those whose lowest T score was on Mf (Black, 1956). The results showed that a Feminine and Androgynous sex role type on the BSRI was more common in the "feminine" Mf groups and that a Masculine sex role type was more common in the "masculine" Mf groups, using both Mf classification methods. Overall, for the entire range of Mf scores asser- tiveness and Mf scores were not significantly correlated. When the extreme groups on Mf were analyzed for differences in assertiveness the results differed for the two Mf classi- fication methods: Using quartiles, the Lowest and Highest Quartiles on Mf did not differ in assertiveness; for the Highest Point-Lowest Point method, however, the mean asser— tiveness score was significantly higher in the Highest Point ("masculine") group than in the Lowest Point ("feminine") group. Within the low Mf score groups, both the Lowest Quar- tile and the Lowest Point group, assertiveness scores were clearly related to BSRI sex role type. The Masculine BSRI Nancy Jeanne Egan group was most assertive, followed by the Androgynous, Feminine, and Undifferentiated groups. This relationship between assertiveness and BSRI category also held for the entire sample. These findings indicate that when Mf is the lowest profile point the individual may be low on assertiveness, but that it is BSRI category rather than Mf score which is most clearly associated with assertiveness. The results argue for caution in interpreting the meaning of low Mf T scores for women. A considerable number of low Mf scorers appear to be interpersonally assertive rather than passive and submissive. Additionally, the discussion includes in- formation on the MMPI profile patterns of this sample in contrast to a 1970 study. Methodological difficulties and suggestions for future research are also discussed. ACKNOWLEDGMENTS There are a number of people who have helped me with this research and deserve my thanks. I especially want to thank Dr. Al Aniskiewicz, my dissertation chairperson, for his enthusiastic support, his thoughtful questions, and helpful guidance throughout all phases of this project. From my beginning days of practicum onward his knowledge, respect, and encouragement have invaluably influenced my developing professional skills and confidence. Dr. Lucy Ferguson, my doctoral guidance committee chairperson deserves my thanks for her consistently sound and reasonable advice. Her guidance in research and the planning of my graduate program has provided direction while allowing for independent development of my professional goals and interests. I would like to express appreciation to Dr. Gil DeRath for his thought-provoking comments, his thorough editorial assistance, and his respectful challenges. I am grateful to Dr. William Crano for his assistance in design and analysis as well as for his more specific suggestions regarding statistics. My friends have been an important support throughout my graduate career and my work on this dissertation. I want to thank Wendy Fielder for her continued friendship and thoughtfulness since our first overwhelming days of ii graduate school; David Kahler for the emotional sustenance and his unwavering confidence in my abilities; Pat Ponto for the way she listens and for openly sharing my joy at each stage of accomplishment; and Michelle Klee for the lively discussions about femininity and personality develop- ment. I am also grateful to Richard Genirberg for his patient and cheerful help in negotiating with the computer which made that part of the work much easier. And during the final stages of this project I very much appreciated Micheal Sherry's sense of humor, perspective, and under- standing. iii TABLE OF CONTENTS Chapter I INTRODUCTION . . . . . . . Construction and Development of Scale 5 (Mf) . . . . Content of the Mf Scale Items . . Research on Mf in College Women The Extreme Groups Method . . . Analysis by Relative Elevation. . Bipolarity and Unidimensionality of the Mf Scale . . . The Concept of Psychological Androgyny. . . Other Measures of Psychological Androgyny. . Validity Studies and Other Research on the BSRI . . . Androgyny in Relation to Psycho- logical Health and Adjustment. Assertiveness and the College Self—Expression Scale (CSES) . . Construct and Concurrent Validity of the CSES . . . . . . Behavioral Criterion Validation of the CSES . . . . . II HYPOTHESES. . . . . . . Hypotheses Related to Previous Research Findings and Methodology. Hypotheses Related to Clinical Assumptions . . . . . . III METHOD. . . . . . . . . Subjects . . . . . . Materials and Procedures . . . Scoring of the Measures . . . MMPI . . . . . The Bem Sex-Role Inventory. . The College Self— -Expression Scale. . . . . . iv 17 19 21 23 28 30 32 35 35 38 41 41 41 42 42 42 43 Chapter IV RESULTS . . . . . . . . Results of Hypothesis Testing. . Hypothesis 1A. . . . . Hypothesis lB. . . . . Hypothesis 1C. . . . . Hypothesis 2 . . . . . Hypothesis 3 . . . . Hypotheses 4A through 4C . . . Hypothesis 4A. . . . . Hypothesis 4B. . . . . Hypothesis 4C. . . . . Hypothesis 5 and 6 . . . V DISCUSSION. . . . . . . . APPENDICES. q l REFERENCES. til U 0 CD I The MMPI Profile Patterns--A Socio-Cultural Perspective . . Comments on the Bem Sex Role Inventory Scores. . . . Results Associating Masculinity- Femininity on the Mf Scale to Bem Sex Role Inventory Categories Hypotheses Relating Mf Scores to Assertiveness on the CSES . . The Lowest Quartile Mf Women--The Importance of BSRI Category in Relationship to Assertiveness . Methodological Criticisms and Implications for Future Research. DESCRIPTIVE STATISTICS OF THE MF SCALE SCORE GROUPS. . . . HIGH POINT AND CLINICALLY SIGNIFICANT PROFILE CODES. . . THE BEM SEX ROLE INVENTORY AND SCORING KEY . . ANALYSIS OF VARIANCE SUMMARY TABLES CSES AND BSRI STATISTICS FOR THE MF SCALE GROUPS . . . THE COLLEGE SELF-EXPRESSION SCALE . Reference Note. . . . . . . Page 45 47 47 49 50 51 51 52 52 55 56 58 60 60 63 64 69 71 75 78 78 79 81 83 84 86 90 97 Table 1. LIST OF TABLES Page Raw Score Means with K Correction and T Score Equivalents for College Women (N = 419) on the Validity and Clinical Scales of the MMPI . . . . . . . 45 Number of Participants in Each Category of the Bem Sex Role Inventory . . 47 Frequency of Subjects in the Lowest and Highest Mf Quartiles Categorized as Masculine or Androgynous and as Feminine on the Bem Sex Role Inventory . . . . 48 Mean Assertiveness Scores on the College Self-Expression Scale for the Lowest and Highest Mf Quartile Groups . . . . . 50 Number of Subjects Classified as Androgynous or Masculine and as Feminine on the BSRI for the High Point 5 and the Low Point 5 Groups of the MMPI . . . . 53 Means of Bem Femininity and Masculinity Scale Scores for the High Point and Low Point Mf Scale Groups . . . . . . 54 Mean Assertiveness Scores on the CSES for the High Point 5 and the Low Point 5 Groups of the MMPI. . . . . . . 55 Mean Assertiveness Scores on the College Self-Expression Scale of Androgynous and Feminine Groups whose Lowest MMPI Profile Point is Scale 5 (Mf) . . . . . . 56 Mean Assertiveness Scores on the College Self-Expression Scale for the Bem Sex Role Inventory Categories . . . . . 57 Descriptive Statistics for all Quartile Groups of Mf Scores . . . . . . . 78 Descriptive Statistics for Groups with Scale 5 (Mf) as the Lowest or Highest Profile Point . . . . . . . . 78 vi Page Percentage of Codes from College Women (N = 419) in Which Each Pair of High Points Occurs . . . . . . 79 Highest Point and Clinically Significant Profile Frequencies from College Women (N = 419) . . . . . . . . . 80 Analysis of Variance Relating Bem Scales to MMPI High and Low Point Mf Groups . . . . . . . . . 83 Analysis of Variance Relating Bem Sex Role Inventory Categories to College Self-Expression Scores . . . . 83 Lowest Mf Quartile Statistics (Feminine) . . . . . . . . . 84 Highest Mf Quartile Statistics (Masculine) . . . . . . . . . 84 Lowest Point Mf Group Statistics (Feminine) . . . . . . . . . 85 Highest Point Mf Group Statistics (Masculine) . . . . . . . . . 85 vii CHAPTER I INTRODUCTION The clinician attempting to meaningfully integrate a woman's MMPI Mf score (Scale 5), into the profile configura- tion will have no trouble finding interpretive suggestions in the more widely used MMPI manuals. On the other hand, many research studies have not included the Mf Scale and clinicians have often been unclear as to the meaning of the scale. Carson (1969, pp. 483-484) states: For females, high scorers (Scale 5) tend in general to be aggressive, dominating and competitive; they are found in large numbers in activities and occupations that are traditionally male. . . . Low 5 females are passive, submissive, yielding, and demure. . . . Women who achieve extremely low T-scores are usually highly constricted, self-pitying and fault-finding; they seem unable to tolerate pleasant experiences. LaChar's observations are similar, although in his statements he differentiates between normal and psychiatric populations. High scorers in a normal pOpulation are described as adven- turous, confident, and competitive. In female psychiatric patients Scale 5 elevation suggests a rebellious, dominating, competitive individual. Low scorers among normals are con- sidered to be sensitive and modest with typically feminine interests. In addition, LaChar adds that they "may display a masochistic acceptance of discomfort" (p. 8). Women psy- chiatric patients with very low Mf scores are often described 1 2 as highly constricted and very difficult, manipulative clients in therapy (LaChar, 1974). Although neither Carson nor LaChar intend to convey the idea that a woman's Mf score can be considered out of the context of the entire profile, the interpretive state— ments concerning both high and low scores on Mf are strong- ly expressed and based on clinical experience rather than empirical investigation. When interpreting the Mf Scale score it should be remembered that the construction and development of this scale differed from that of the other scales. Also, surprisingly little research has been done on it, particularly with women (Dahlstrom, Welsch, & Dahl- strom, 1972). Construction and Development of Scale 5 (Mf) Scale 5 was originally developed to identify person- ality features associated with male sexual inversion. The scale consisted of an initial item pool and items added later from the sexual inversion studies of Terman and Miles (1936). As a result, the response frequencies of the orig- inal Minnesota criterion group were not available, requiring the use of special groups of normals--54 soldiers and 67 female airline employees (Hathaway & McKinley, 1940). The published scale was derived by contrasting the normal men with 13 carefully selected male sexual inverts. Comparisons were also made between more feminine but otherwise normal males (as determined from the Terman and Miles I scale), and 3 more typical males. As a final, less important criterion the items were also compared for male-female differences because the dimension of inversion being examined appeared in the studies of Terman and Miles (1936, 1980) to be psy- chologically similar to differences between men and women (Hathaway, 1956). A criterion group of women was utilized in the devel- Opment of the Mf scale but clearly the purpose of the scale was not to aid in the diagnosis or understanding of women's personalities. There was one attempt to improve the Mf scale by developing a corresponding Fm scale to identify female sexual inverts. Records of women whose problems included homosexuality were used as the criterion group but the scale proved to be inadequate upon cross-validation (Hathaway, 1956). The Mf scale as it is currently in use consists of 60 items, 23 of which were taken from the Terman and Miles scale and 37 from the original MMPI item pool. The content is heterogenous, ranging from occupational preferences, social activities and hobbies, to fears, personal sensitiv- ities, and sexual practices (Dahlstrom, et al., 1972). It is generally acknowledged that it is not a pure measure of masculinity-femininity; however, questions about what it does measure, especially in women, remain unanswered. Despite controversy, it is a highly reliable scale. The items have the format, content, and response direction 4 associated with response stability, although some shifts have been noted in treated groups (Schofield 1950, 1953). Content of the Mf Scale Items Criticism of the content of the Mf scale items has previously faced the counterargument that the empirical approach to test construction is not interested in content per se, but rather in response differences between criterion groups. Although this may be a valid argument for the items on the clinical scales, it does not hold for the concept of masculinity-femininity, especially as it is so heterogene- ously measured by Scale 5. For women, a high Mf score implies that she is answering in the direction opposite to a group of male sexual inverts. In the development of the scale, a criterion group of sexually abnormal women was not successfully included; to compare the scores of the 67 women airline employees in the 1940's to women's responses today without the prior criterion group validation makes any strong interpretive statements highly suspect. Constantinople (1973) in her discussion of some common assumptions regarding masculinity-femininity ques- tions whether the construct is best defined in terms of sex differences in item responses. The strictly empirical ap- proach to test construction would include any item that dis— criminates men from women regardless of the triviality or cultural-historical specificity of the item. "In all prob- ability the length of the big toe would discriminate men 5 from women but does having a longer big toe than most women make a woman less 'feminine?'" (Constantinople, p. 405). Also, on the Mf scale the meaning of some of the explicitly sexual items may have changed over time, especially the ones which are reversed scored for men and women. For example, a woman's False response to "I wish I were not bothered by thoughts about sex" is scored in the feminine direction. Perhaps a very "feminine" woman does not con- sciously experience many explicitly sexual thoughts and is, therefore, not bothered by them. An androgynous woman may give a similar reply for an entirely different reason--she may actively enjoy her explicitly sexual thoughts and fanta- sies. Changes in societal mores and sexual behavior may likewise have tampered with the reverse scoring and meaning of such items as "I like to talk about sex" and "I have never indulged in any unusual sex practices." Although the Mf scale was constructed using an empir— ical approach, recall that the male-female sexual differen- tiation of the items was not the primary basis for selection. Drake (1953) developed a Sex Differential scale using the MMPI items that showed the largest differences in frequency of endorsement for males and females. Despite considerable overlap with the Mf items, the scales are by no means identical. Drake's scale is based solely on sensitivity to male-female differences-—a dubious conceptualization of masculinity-femininity according to Constantinople, 6 nevertheless, its empirical base is clear. The Mf scale, on the other hand, attempts to assess male-female differ- ences as well as normal male-invert male differences and combine them in one scale. It seems only reasonable that the validity of this measure would be greater for men than for women; yet, there is no dearth of interpretive hypothes- es for "deviant" Mf scores in women, especially low ones. Carson (1969) in describing the content of the Mf scale states that the items have to do with "interests, vocational choices, aesthetic preferences and an activity- passivity dimension" (p. 482). Most of the clinical inter- pretive suggestions heavily emphasize the activity-passivity dimension (e.g., passive, submissive, yielding, and aggres- sive, dominating, competitive); however, upon examination the items show this dimension to be directly assessed in very few questions. Pepper and Strong (1958) rationally divided the Mf scale into 5 subscales of content as follows: Personal and Emotional Sensitivity, Sexual Identification, Altruism, Feminine Occupational Identification, and Denial of Masculine Occupations. There is no specific mention of an activity-passivity dimension in Pepper and Strong's con— tent typology, although the 9 items in the Altruism and the 15 items in the Personal and Emotional Sensitivity category include questions suggestive of such a dimension. Some individuals who respond in a given way to these items, may, indeed, be passive and submissive. Possibly others who 7 respond similarly may be sensitive and altruistic in a more active, efficacious way. Furthermore, most of the items on the Mf scale deal with interests, occupations, and preferred activities. Ac- cording to Pepper and Strong's content divisions, a total of 27 items concern feminine occupations or denial of masculine occupations. Certainly preferring to read love stories rath- er than go hunting represents a physically passive form of enjoyment, but this type of passivity need not extend to the interpersonal, psychological, or intellectual realms. For college students, the masculine and feminine interest items are best at differentiating males from females (Little, 1949; Murray, 1963). Both male and female college students score higher in the feminine direction than the original Minnesota sample of men (Bechtoldt & Dahlstrom, 1953). It is the items dealing with emotional sensitivity, altruism, and aes- thetic interests which largely account for the similarly high femininity scores of college men and women. In fact, in this study Scale 5 was the most frequent peak in college men with 30% having 5 as the first or second highest score. There is still a question as to whether college students are preselected for greater femininity or move in this direction as a result of the college experience. Most of the evidence at this time favors the selection hypothesis (Brehm, 1954; Mills, 1954). More recently Kokosh (1969) found that students in the upper range of GPA are higher on Mf than those with 8 lower grades. Barger and Hall (1964) found that both males and females who score in the masculine direction on the Mf scale drop out of college at a higher rate than those who score in the feminine direction. Femininity as measured by the MMPI scales does not seem to interfere with academic achievement. Research on Mf in College Women Despite the tremendous number of MMPIs which have been administered to college students in the name of scien- tific study, the researchers agree that little definitive work has been done on the correlates of low 5 and high 5 profiles in women (Dahlstrom, et al., 1972; Klopfer, 1966; LaChar, 1974). The three major studies on Mf in college women were all conducted in the 1950's and each utilized slightly different research methodology (Black, 1956, (a), (b); Hathaway & Meehl, 1952). The Extreme Groups Method In the Hathaway and Meehl study (1952) the extreme scores method was used, e.g., the responses to an adjective checklist of those who scored in the top quartile on Mf were compared to those who scored in the lowest quartile. Each subject checked adjectives to describe themselves. Ratings of the subject by friends and acquaintances were also ob- tained. The advantages of using an adjective checklist include systematic coverage and directly comparable results 9 for all individuals and scales; however, this method does not allow for quantitative distinctions, qualifications, or standardized interpretation of the adjectives to be applied. The extreme scores methodology has a number of weak- nesses which need to be considered when interpreting the re- sult of Hathaway and Meehl's work. Individuals in the same quartile for comparison on Scale 5 may differ greatly on the rest of the scales. Moreover, for some individuals other parts of the profile may be higher or lower than the scale score being considered, i.e., those in the lowest quartile on Scale 5 will have a variety of high points as well as a variety of low points. Dahlstrom, et a1. (1972) point out that although, in general, the variation of high and low points in the quartile may cancel out, there are high cor- relations between scales in some populations. This may lead to contaminated results: Criterion groups formed using the extreme groups method may show characteristics from several MMPI variables because they are simultaneously high on several correlated scales. The actual results of Hathaway and Meehl's research on Scale 5 are worth examining, keeping in mind the limita- tions imposed by the methodology. Dahlstrom et a1. (1972), extensively summarized the results of this study as follows: When females high on Scale 5 (the masculine direction), were rated by peers the only adjective checked significantly more frequently than for the other groups was "adventurous." 10 The self-descriptions included adventurous as well as "having physical strength and endurance, poised, easy-going, relaxed, balanced, logical and facing life." Those low on Scale 5 (the feminine direction), were described by others as "sensitive, responsive, modest, grateful and wise"; these women saw themselves as "sensitive and idealistic." Low 5 scores for women are much more common than high scores but little research has been done on the correlates of either type. Hathaway and Meehl comment: "It is a clinical impres- sion that low Mf scores in females represent an almost maso- chistic passivity" (Black 1956a, p. 161); however, their research does not support such an extreme statement. Analysis by Relative Elevation Black's research (1956a) uses a slightly different methodology which may account for discrepancies between his findings and those of Hathaway and Meehl. In the Black study of adjectives associated with the various MMPI codes, criterion groups were formed on the basis of relative eleva- tion within each individual's profile. The adjective check- lists of women whose highest score was on Scale 5, and those whose lowest score was on 5 were compared with the lists from the total pOpulation, minus the subgroups under con— sideration. This method gives greater homogeneity to the criterion groups but does not obviate all difficulties, e.g., the peak score of a woman whose highest score is on 5 may not be reliably higher than her other scores; upon repeated ll testing such a profile could possibly be categorized in another criterion group. Also, cases in each group may differ in the absolute level of the peak score by as much as three or four standard deviations (Dahlstrom et al., 1972). One way to improve this method is to categorize profiles by the peak score or scores while also taking into considera- tion the absolute elevation. The high-point 5 females in Black's adjective study were significantly more often described by peers as "indeci- sive, rebellious, natural, and unrealistic," than were other groups. They were less frequently seen as "poised, dreamy, polished, or sensitive." In their self-descriptions, those whose peak was on Scale 5 described themselves as "rough, incoherent, shiftless, and unemotional." Infrequently checked self-descriptions of this group included "popular, good-tempered, polished, peaceable, kind, lively, alert, sentimental, and emotional"; they did not consider themselves as having wide or aesthetic interests. Those women whose lowest point in the profile was on Scale 5 were described by their peers as "worldly, popular, decisive, and versatile." The adjectives "energetic, undependable, shy, rough, unreal- istic and disorderly" were more infrequently checked by the peers of the low 5 females than by peers of the rest of the sample. The low 5 women paint a less flattering self-por- trait than the peer descriptions indicate: They describe themselves as "self—distrusting, self-dissatisfied, moody, 12 polished, shy, sensitive, neurotic, unrealistic, talkative, sentimental," and as having aesthetic interests. Self- portraits by the low 5 female infrequently included "bal- anced, independent, decisive, good-tempered, practical, relaxed, and modest." Black makes several interesting points in discussing the results of his research. He notes that while the peer descriptions do not provide a detailed or clearcut picture of the low 5 scorers, a picture of masochistic passivity is certainly not in evidence. "Energetic" is infrequently checked, yet this is not a sufficiently strong indication of a submissive, passive stance, and the other adjectives por- tray an effectively functioning personality. The self- descriptions do, on the other hand, suggest a person who does not see herself positively or as effectively defended against her shortcomings (Black, 1956a). On the basis of his findings for the high 5 women, Black asserts that the adjectives do not support the char- acterization of "driving, vigorous, and determined" as set forth by Gough (1952), nor is there evidence that it points to "abnormal sexual interest" (Fry, 1949). Perhaps these women may have more masculine interests which result in their being viewed as unrealistic, confused, and rebellious. The self-descriptions are quite different from those of their peers but may be indicative of rebellion against societal stereotypes (incoherent, shiftless, unemotional, 13 rough, deliberate). Both of the studies of Hathaway and Meehl (1952) and Black (1956a) were conducted in the 1950's before the bur- geoning growth of the Women's Movement. The high 5 profile which was infrequent among college women of the 50's may be somewhat more common today. Furthermore, this research did not consider the other scale elevations in relation to the scores on 5. For some configurations Scale 5 may indeed serve as a measure of dominance-submission, whereas for other patterns it may function differently. Dahlstrom, et al., (1972) comments: It is observed clinically that the higher the elevation of the neurotic triad, the lower the value of scale 5 will be in this group. This configuration frequently ac- companies a masochistic trend in the adjus- tive efforts of the woman with self-depre- ciation, long-suffering sacrifice, and unnecessary assumptions of burdens and responsibilities (p. 195). Other configurations with a low 5 might not exhibit such masochistic or submissive behavior patterns. The results of another study by Black (1956b) offer support for the idea that Scale 5 may function differently for various profile configurations. In this study, female college students nominated their peers into 15 personality categories such as most likeable, shy, career-minded, inde- pendent, naive, doesn't fit in, best leader, etc. The three or four women in each residence hall receiving a sub- stantial plurality of votes were chosen to be in each 14 criterion group, with a total of 12-25 in each of the 15 groups. Although this is an intriguing study which uses a normal p0pu1ation and meaningful criteria, the profiles were analyzed by considering the mean scale scores of each cri- terion group rather than the profile configurations of individuals. It should be kept in mind that the profiles for each of the 15 groups are not those of any individual, and different scales may have more variation than others. For several of the 15 categories interesting and significant results were obtained on Scale 5: In the group voted "most independent" there was a strikingly large stand- ared deviation on almost all scales. On Scale 5 there was a small cluster of women with scores above 70 T, another cluster around 45 T, and only one score between 49 and 70 T. The bimodal distribution on 5 for this group indicates that both those with very masculine and those with very feminine interests, as measured by the MMPI, were considered to be "independent." The group voted "most sensitive" also had a bimodal distribution on Scale 5. Over one-half of the sensitive group had very low Scale 5 scores and the rest were above the sample mean; seven individuals were in the 37-41 T range and five fell between 51 and 57 T. Black hypothesizes that there may be two types of individuals considered to be hyper- sensitive--those who express it intropunitively via depres- sion, withdrawal and anxiety, and those whose sensitivity 15 is manifested extrapunitively through defiant, tactless or vindictive behavior. Neither the most sensitive nor the most independent group display a typical profile configura- tion to go along with the bimodal distribution on 5; how- ever, the fact that low and high 53 are found in both of these very different categories supports the idea that not all low or high women will display the same type of active- passive or dominant-submissive behavior, irrespective of the rest of the profile. Another example of the intricacies involved in ana- lyzing profiles by individual high points or by using ex— treme groups is illustrated by Black's findings concerning Scale 6 (Paranoia). A high score on Scale 6 (near 70 T) tended to be associated with shyness in the college women studied whereas a pggk on 6 tended to be associated with leadership. None of the leaders scored higher than 65 T and 1/3 of the shy group scored 67 T or higher, while only 5.5% of the shy group had a pgak on 6. A modest elevation on 6, especially if it is the peak score, may indicate leadership potentialities in college women but extreme ele- vation may be related to shy behavior. It is likely that Scale 5 is also more complex than commonly thought. The generally accepted interpretive statements concerning high and low Scale 5 scores need to be modified accordingly. 16 Bipolarity and Unidimensionality of the Mf Scale The often unclear results and bimodal distributions obtained in the Mf Scale research may be due to the assump- tions of bipolarity and unidimensionality of the masculinity- femininity construct which underlie the scale. Constantin- ople (1973) questions both the assumption that masculinity- femininity is a simple bipolar dimension with masculinity at one end and femininity at the other, and that it can be adequately described by a single score. In both a factor analytic study and in a rationally-derived categorization, the masculine and feminine interest items of Scale 5 were found to be in separate categories rather than opposite ends of a single bipolar continuum (Graham, Schroeder & Lilly, 1971; Pepper & Strong, 1958). Constantinople argues that the M-F construct is more accurately represented by several scores, one for each of the various subtraits under the masculinity-femininity di- mension as presently understood. Substantiation for the usefulness of the multidimensional approach is found in an early report by Webster (1956) on developmental changes in M-F for Vassar students. Items known to discriminate the sexes were divided into three content subscales: I. Con- ventionality (preferences for conventionally feminine inter- ests); II. Passivity (docility, modesty, lack of aggressive- ness, manipulativeness); and III. Feminine sensitivity (emotionality, fantasy, "neurotic trends," aesthetic 17 interests). Although freshman-senior differences were small, seniors were significantly higher than freshmen on III and lower on I and II. Webster concluded that these college women became more masculine in terms of less con- ventional, passive behaviors and more feminine with regard to a greater awareness of inner life. It seems likely that the Vassar seniors had an expanded repertoire of behaviors which did not preclude them from scoring high in the femi— nine direction of the scale. The Concgpt of Psychological Androgyny The concept of psychological androgyny (Bem, 1974), is a more recent formulation of the masculinity-femininity construct which takes account of Constantinople's criticisms and research findings such as those of Webster (1956). According to Bem, the psychologically androgynous person has both masculine and feminine behaviors in his/her reper- toire. ‘Such an individual is able to display those that are appropriate to the situation without reliance on stereo- typic sex role proscriptions. Androgyny further implies that an individual may even blend these com- plementary modalities into a single act, being able, for example, to fire an employee if the circumstances warrant it, but do it with sensitivity for the human emotion that such an act inevitably pro- duces (Bem, Martyna, & Watson 1976, p. 1016). 18 The Bem Sex—Role Inventory (Bem, 1974) was designed to measure masculinity and femininity as two independent dimensions. It is a self-report inventory consisting of 60 adjectives, 20 of which were previously rated by college students as typically feminine, 20 as typically masculine, and 20 as neutral. These are rated on a seven-point scale ranging from "always or almost always true" to "never or almost never true"; items were chosen on the basis of sex- typed social desirability, not actual differential endorse- ment by males and females. In Bem's original scoring scheme a person's androgyny score was based on the differ- ence between the Masculinity and Femininity Scale totals. The closer the score was to zero, the more androgynous the individual. However, this scoring method received much criticism because it did not differentiate those who scored high on both Masculinity and Femininity Scales from those who scored low on both (Bem, et al., 1976; Bem, 1977; Spence, Helmreich & Stapp, 1975; Strahan, 1975). Subse- quent research on the differences between the low Masculin— ity-low Femininity and the high Masculinity-high Femininity groups has supported the distinction between them: Signifi- cant differences between the two groups have not been found on all variables studied but low-low scorers do appear to be lower in self-esteem and to display lesser amounts of some androgynous behaviors studied in the laboratory (e.g., responsiveness to a kitten, self-disclosure), than do those 19 scoring high on both the Masculinity and Femininity Scales (Bem, 1975; Bem, et al., 1976; Schiff, 1977; Spence, et al., 1975). The revised scoring method makes use of the four-fold classification method suggested by Spence, et al., (1975): Those individuals who score above the sample medians on both the Masculinity and Femininity Scales are considered to be androgynous and those who score below both sample medians are "undifferentiated." If the Femininity score is below the median and the Masculinity score above, the individual is classified "masculine"; one whose Femininity score is above the median and Masculinity score below is classified "feminine." Other Measures of Psychological Androgyny Bem's concept of androgyny has stimulated a number of researchers to devise similar measures, although the BSRI has received the most research attention and validation (Kelly & Worell, 1977). Spence, et a1. (1974, 1975) develop- ed the Personal Attributes Questionnaire using items from the Sex Role Stereotype Questionnaire of Rosenkrantz, Vogel, Bee, Broverman & Broverman (1968). It consists of three scales (male-valued, female-valued and sex-specific), which were developed by asking college students to rate character- istics "more typically found" in one sex than in the other. Although the content of the gender-valued scales is very similar to that of the BSRI, the FAQ items were judged to be 20 ideal for both men and women (but more typical of one sex), whereas the BSRI items were chosen for sex-based social desirability. The PRF ANDRO is a scale derived from the previously standardized Personality Research Form (Jackson, 1967). Items were initially selected from the PRF pool on a "rational-intuitive" basis and later statistically refined. The respondent answers true or false to each self-descrip- tive statement rather than rating items on a 7-point scale. In order to control for acquiescent response distortion, items are keyed in both the positive and negative direction (Berzins, Welling & Wetter, 1975). Heilbrun (1976) also constructed independent mascu- linity and femininity adjective scales similar to the BSRI, FAQ, and PRF ANDRO, although there are several construction- al and psychometric differences. Included are items from the Adjective Checklist "that discriminated between college males identified with masculine fathers and college females identified with feminine mothers" (Heilbrun, 1976, p. 184). This scale combines a measure of psychological masculine- feminine identification with gender differences in item endorsement. Both socially desirable and undesirable traits are included but the effect of this mixture has not yet been assessed (Kelly & Worell, 1977). The sex role inventories presently used to assess androgyny have varying degrees of differences in psychometric 21 and constructional properties and in item selection proce- dures. Thus far there has not been sufficient comparative research on these instruments to determine whether they are defining sex-typed, androgynous, and undifferentiated roles in the same way. One study compared the BSRI and the PRF ANDRO scores of college students by examining the agreement in the assignment of sex role category. For both sexes combined only 77 of the 158 respondents (42%) fell in the same sex role category (Gayton, Havu, Ozmon & Tavormina, 1977). Because there may be conceptual differences between the various androgyny measures that are presently indeter- minate, the proposed research will utilize the BSRI which has been most extensively validated. Validity Studies and Other Research on the BSRI Bem has conducted a number of laboratory studies on the BSRI in order to behaviorally validate the self-reported sex role orientation measured by the instrument. Her re- search has indicated that an androgynous individual of either sex is able to display "masculine" independence when under pressure to conform, in addition to "feminine" nurtur- ance when interacting with a kitten. Sex-typed individuals on the other hand, were low in one or both of these behaviors (Bem, et al., 1976). Masculine females were quite independ- ent but they were also pp: significantly less nurturant toward the kitten than were androgynous females. In two other situations, however, the masculine women were less 22 nurturant than androgynous women (Bem et al., 1976). A lack of difference between the masculine and androgynous women was also found on a measure of self-esteem employed by Schiff (1977). Her study revealed a significantly greater degree of self-esteem among androgynous women than among feminine or undifferentiated women, but no difference in self-esteem between androgynous and masculine women. It appears thus far that on some variables and in some situa- tions the distinction between androgynous and masculine women, as measured by the BSRI, is not as clear-cut as would be theoretically expected (Haglund, 1978). For males, Bem's behavioral laboratory studies clear- ly and consistently validate the BSRI. Only androgynous males were high in both the instrumental and expressive domains; that is only androgynous males were found to stand firm in their opinion as well as to cuddle kittens, bounce babies, and offer a sympathetic ear to someone in distress (Bem, et a1. 1976, p. 1022). The results for women, as is also true on most conventional measures of M-F, are less consistent but show the same gen- eral pattern. Masculine individuals of either sex are high in independence but low in nurturance whereas feminine indi- viduals are high in nurturance and low in independence. The androgynous individuals of both sexes are capable of nurtur- ance and independence, both masculine and feminine behaviors, thus supporting the idea that the androgynous individual has a less restricted behavioral repertoire (Bem, et al., 1976). 23 Although the validity research on the Bem Sex-Role Inventory has not supported it unqualifiedly, many of the critical studies had not used the revised method of scoring (Hogan, 1977; Strahan, 1975; Spence, et a1, 1975). Factor analytic studies have supported Bem‘s contention that the Masculinity and Femininity scales of the BSRI measure inde- pendent constructs rather than a single bipolar trait; how- ever, several item changes and deletions are recommended to increase the homogeneity and interpretability of the scales (Gaudreau, 1977; Waters,Waters & Pincus, 1977; Whetton & Swindells, 1977). In two studies the items "Masculine," "Feminine," and "athletic" were found to do little except identify the biological sex of the respondent and might well be deleted (Gaudreau, 1977; Waters, et al., 1977). Several additional items were considered weak in some studies but not in others. Given these findings it is apparent that more research is needed before recommending further changes in the scale. Androgyny in Relation to Psychological Health and Adjustment Aside from the factorial validation of the BSRI, there have been attempts at convergent validation of the assumption that the androgynous individual represents "a new and more human standard of psychological health" (Bem, 1975, p. 643). The traditional view of gender role orien- tation has considered a psychologically healthy male to be 24 "masculine" and a psychologically healthy female to be "feminine." Bem's conceptualization of psychological an- drogyny questions this belief. The androgynous person has a less restricted range of behaviors and should be able to engage in more adaptive, situation-appropriate behavior un- hampered by sex-role stereotypes (Bem, 1974). Research prior to Bem's work offers some indirect support of the idea that a high level of sex-typing may not be the most psychologically adaptive gender role orientation. High femininity in females has been associated with low self-esteem and high anxiety (Cosentino & Heilbrun, 1964; Sears, 1970; Webb, 1963); high masculinity in adult males has been associated with high anxiety, neuroticism, and low self-acceptance, despite better psychological adjustment among masculine males during adolescence (Harford, Willis & Deabler, 1967; Mussen, 1962). Yet, an androgynous person may encounter conflicts in a society which still expects a degree of culturally-approved sex role behavior from its members. Haglund (1978) in a study of college students examin- ed the adjustment variables of achieving tendency, social self-esteem, role consistency, sensitivity to rejection, and manifest anxiety in relation to androgyny as measured by the BSRI. In general, the androgynous respondent showed better adjustment and less maladjustment on these variables than did the feminine respondents. On some variables, how- ever, the difference between masculine and androgynous 25 individuals was nonsignificant. As previously mentioned, Schiff (1977) also found androgynous women to be higher on self-esteem than feminine and undifferentiated women but no difference between the masculine and androgynous individuals. This may reflect the scale's inability to differentiate mas- culine and androgynous orientations, but possibly it reflects the more adaptive stance generally associated with a mascu- line orientation, e.g., achievement, confidence, assertive- ness, high self—esteem. Consider once again the interpretive statements in the MMPI manuals that are associated with high and low Mf scale scores: low scorers are passive, submis— sive, masochistically self-sacrificing, etc.; high scorers are adventurous, driving, competitive. Although neither characterization is unduly flattering, it seems likely that the latter individual would score higher on the adjustment variables employed in the Haglund study. The lack of differentiation between the androgynous and masculine women on some psychological adjustment vari— ables is congruent with the belief that cross-sex typing is less maladaptive for women than it is for men. Masculine characteristics are considered as more representative of a psychologically healthy individual and women are given more latitude in terms of sex role deviancy (Broverman, Vogel, Broverman, Clarkson & Rosenkrantz, 1972; Mischel & Mischel, 1971). LaTorre (1976) found that gender role problems are more evident in male than female schizophrenics. Similarly, 26 a study by Heilbrun (1973) revealed that the relationship between gender role and psychopathology was reliable for males but not for females. These findings are consistent with LaTorre's recent work (1978) comparing college students' scores on the BSRI with psychological adjustment as measured by Lanyon's Alienation (A1) scale and Eysenck's Neuroticism (N) scale (Eysenck & Eysenck, 1968; Lanyon, 1973). Overall, gender role was more related to the psychological adjustment of males than of females. The androgynous males obtained better adjustment scores than those who were classified masculine, feminine, or undifferentiated. Once again, for females the results were less clear-cut, but the one signi- ficant and consistent finding was that feminine individuals of both sexes obtained the highest maladjustment scores. Feminine females were notably higher on neuroticism than those classified masculine or androgynous (LaTorre, 1978). Studies by Heilbrun (1976) and Nevill (1977), using several different types of measures, also found androgynous individ- uals to show better adjustment. The evidence relating femininity as measured by the BSRI to psychological maladjustment appears congruent with the interpretations given to low Scale 5 scores on the MMPI. But an important difference between the two scales must be remembered: The MMPI measures masculinity-femininity as a bipolar trait and does not allow for the determination of an individual's degree of androgyny. A middle range score on 27 the Mf Scale would not necessarily indicate an androgynous individual. It is proposed that of the college women who score low on Mf a considerable percentage are, in fact, an- drogynous. The clinical observation that the higher the neurotic triad, the lower the Scale 5 score (Dahlstrom, et al., 1972) may also be empirically supported because not all women with low 5s will have a high neurotic triad elevation. The Mf Scale includes several aspects of femininity, but the final score does not differentiate among them. An androgynous woman and a highly feminine woman might share a considerable number of "typically feminine" interests and fantasies, such as raising houseplants or wanting to be a singer, using two items from the MMPI. It is also highly probable that they both played with dolls at one time, which is another item from the MMPI. Because more of the scale items deal with interests and vocations than with emotional hypersensitivity, "neuroticism," or passivity, possibly two very different women in terms of interpersonal style and psychological adjustment could obtain similar Mf scores. Clinicians, of course, may derive information about these areas from other clinical scales. The question of whether "femininity" is or is not an effective disposition for college students further exempli- fies this confusion. LaTorre and Gregoire (1977) and Biller (1973) contend that masculinity is a more effective behavior- al trait for college students because academia is a masculine 28 environment. However, as previously mentioned, a higher GPA and a lower dropout rate is associated with Mf Scale scores higher in the feminine direction (Barger & Hall, 1964; Kokosh, 1969). Webster's findings (1956) on the Vassar women are relevant here also. Sensitivity to feelings and aesthetic interests are the more healthy aspects of feminin- ity conducive to effective functioning, whereas poor self- image and lack of assertiveness are not. Unfortunately, some constructional features of the Mf scale (the assumed bipolarity and unidimensionality, and item heterogeneity weighted in favor of occupations and interests) do not allow for adequate separation of the neurotic type of feminine woman from the one who is more interpersonally androgynous. One way to empirically test this supposition regarding the Mf scale is to compare low 5 women on a measure of inter- personal assertiveness which is independent of feminine interests, hobbies, sexuality, and awareness of their inner emotional life. Assertiveness and the College Self- Expression Scale (CSES) Alberti and Emmons (1970) define assertiveness as behavior which enables a person to act in his/her own best interests, or to stand up for oneself without excessive anxiety, to express one's rights without denying those of others. Conversely, the unassertive person 29 is unable to express . . . true emotional feelings, fails to refuse unreasonable requests and does not stand up for . . . legitimate rights . . . (such a person) is often taken advantage of and according- ly has little sense of self-esteem or personal dignity. The inability to assert one's rights results in smoldering resent- ment and suppressed hostility, which in turn frequently produces inappropriate feelings of guilt and remorse (O'Leary & Wilson, 1975, p. 245). Research has found that lack of assertiveness is related to compliance tendencies (Eisler, Miller & Hersen, 1973), low self-acceptance (Tolor, Kelly & Stebbins, 1976), and diffi- culties in self-image (Rathus, 1975). This description of the unassertive person closely parallels that of the low Scale 5 woman, i.e., submissive, masochistically accepting of discomfort, highly constricted, one who assumes unnecessary burdens and responsibilities, and is self-depreciating. Thus it would be expected that if the low 5 woman possesses these characteristics she would also be low on a measure of assertiveness. The best presently available instrument for measuring assertiveness in college women is the College Self-Express- ion Scale (Galassi, DeLo, Galassi & Bastien, 1974). This 50-item self-report scale taps feelings aroused in a variety of interpersonal contexts: with family, strangers, business contacts, authority figures, and peers of both sexes. Test- retest reliability for two college samples over a two-week interval has been reported as .89 and .90. The College Self- 30 Expression Scale (CSES), assesses three dimensions of asser- tiveness: Positive, Negative and Self-denial. Positive Assertiveness consists of expressing love, admiration, fond- ness, agreement, and acceptance, whereas negative assertive- ness involves the expression of anger, disapproval, annoy- ance, and dissatisfaction. Self-denial consists of over- apologizing, extreme interpersonal anxiety, and excessive concerns about the feelings of others. Respondents indicate agreement or disagreement with 21 positively worded items and 29 negatively worded items using a 5-point Likert scale. The total score is obtained by summing all the positively worded items and reverse scoring, then summing the negative- ly worded items (Galassi, et al., 1974). Construct and Concurrent Validity of the CSES Construct validity was assessed by correlating scores on the CSES with the number of adjectives checked on the various scales of the Gough Adjective Checklist (Gough & Heilbrun, 1965). Assertiveness scores correlated positively, and significantly with the following scales: Defensiveness, Favorable, Self—Confidence, Achievement, Dominance, Intra- ception, Heterosexuality, Exhibition, Autonomy, and Change. The definitions of these scales portray an individual who is spontaneous, expressive, well-defended, confident, and able to lead and influence other people. CSES scores were signif- icantly negatively correlated with Unfavorable Self-Image, Succorance, Abasement, Deference, and the Counseling 31 Readiness Scales. These correlations indicate that low scorers on assertiveness view themselves negatively, tended to be oversolicitous of emotional support from others, and experience extreme interpersonal anxiety. Although the cor- relation between the Personal Adjustment Scale and the CSES scores was positive, it did not reach significance. Poor adjustment among low scorers on the CSES is, nevertheless, indicated by the correlation with the Counseling Readiness and Unfavorable Self-Image Scales. Also as expected, the CSES scores were not significantly correlated with Aggres- sion, Affiliation, Self-Control, Liability, Endurance, or Nurturance (Galassi, et al., 1974). Current validity was initially established by cor- relating teaching supervisor ratings and self—ratings on the CSES. The correlation was significant but low, E = .19, p <1.01. These raters were untrained, however, and had no other information on the subject other than his/her perform- ance as a teacher, which may partially account for the low degree of association (Galassi, et al., 1974). Three additional concurrent validity studies were then conducted to try to raise the rather low validity of the study (Galassi & Galassi, 1974): (1) Dorm residents' CSES self-rated scores were correlated with their floor counselors' assertiveness ratings of them yielding a correla- tion of .33; (2a) It was predicted that male student legis- lators would be higher on the CSES than male students in 32 general, and the latter would in turn be higher than males in "non-peOple - oriented majors." All groups were ordered in the predicted direction and only the legislators differed significantly from the others; (2b) For females it was pre- dicted that the student legislators would be more assertive than college women in general, and the latter would be more assertive than female students in traditionally feminine majors, e.g., home economics. The female legislators and dorm residents were both significantly more assertive on the CSES than were the women in traditionally feminine majors. The legislators did not score significantly higher than the dorm residents; (3) It was further predicted that students not seeking counseling or those seeking only voca- tional-educational counseling would be more assertive than personal adjustment counselees. Those seeking personal adjustment counseling were found to be significantly less assertive than the other students, a difference also noted in previous research on differences between these groups (Galassi & Galassi, 1973; Heilbrun, 1960). Behavioral Criterion Validation of the CSES The College Self-Expression Scale was further inves- tigated by relating the scores of 81 undergraduates to an overt behavioral criterion. Assertiveness was defined as the subjects insistence that they did, in fact, arrive earlier than the experimental confederate, who claimed to have been there first. The participants who scored higher 33 on the CSES were significantly more behaviorally assertive in this situation. In general, females were significantly more assertive toward members of the same than the other sex (Stebbins, Kelly, Tolor & Power, 1977). Another study of the correspondence between behavior- al and self-report measures of assertiveness concluded that the CSES is the instrument of choice for measuring assertive- ness in college students (Burkhart, Green & Harrison, 1979). Contrary to the frequently encountered finding that behav— ioral role-playing and self-report measures do not covary, this study found considerable overlap between the two types of measures. A factorial study of the CSES by Kipper and Jaffe (1978) revealed four main factors instead of the three originally proposed by Galassi, et a1. (1974). The identi- fied factors were willingness to take risks in interpersonal interactions, the ability to communicate feelings, setting rules and rectifying injustices, and the tendency to invoke a self-punitive attitude. Although the content of these factors does not exactly correspond to the Positive Asser- tiveness, Negative Assertiveness, and Self-denial dimensions originally proposed (Galassi, et al., 1974), logically and psychologically there is sufficient similarity to support the validity claims of the CSES (Kipper & Jaffe, 1978). Overall, the research on the College Self-Expression Scale offers evidence of its utility and promise as a diagnostic 34 and research instrument; test-retest reliability and valid- ity in terms of correspondence with external criteria sup- port its continued use (Bonder, 1975; Burkhart et al., 1979; Galassi & Galassi, 1974; Galassi, et al., 1974; Kipper & Jaffe, 1978; Stebbins, et al., 1977). CHAPTER II HYPOTHESES The hypotheses developed in the current study derive from two major sources: (1) previous research and method- ology with the MMPI; and (2) assumptions based on clinical experience with the instrument. Hathaway and Meehl, and Black performed the seminal empirical investigations of the Mf Scale; therefore, the first set of hypotheses (1 - 4) will replicate their methodologies in relation to my hypoth- eses regarding assertiveness on the CSES and Bem Sex Role Inventory scores. However, since many of the statements included in the MMPI interpretive manuals of Carson and LaChar are based on clinical assumptions, an attempt will be made to empirically support one of these assumptions. The assumption to be examined is whether low T scores on Scale 5 are associated with an elevated neurotic triad pro- file (high T scores on Scales 1, 2, and 3). Hypotheses Related to Previous Research FindingsiandTMethodology 1. Based on the adjective checklist findings of Hathaway and Meehl (1950) and replicating their use of the ex- treme groups' technique (Highest and Lowest Quartiles), the following predictions were made: A. It was expected that significantly more of the subjects in the Highest Quartile on Mf (high T 35 36 scores), would be categorized as Androgynous or Masculine on the BSRI than those in the Lowest Quartile (low T scores); however, the range of Mf scores for Androgynous women was expected to be considerable, including some of the Mf scores in the Lowest Quartile. Conversely, it was ex- pected that significantly more of the subjects in the Lowest Quartile on Mf would be catergoriz- ed as Feminine on the BSRI than those in the Highest Quartile. The clinical interpretations of guides of both Carson (1969) and LaChar (1974) characterized the low 5 female in terms that strongly imply a lack of assertiveness: Carson states, "Low 5 females are passive, submissive, yielding, and demure"; LaChar reports that the low 5 female is "sensitive, modest, and shows a masochistic acceptance of discomfort." Therefore, using extreme groups of high and low 5 females, Highest and Lowest Quar— tiles, it was predicted that there would be a sig- nificant difference in mean assertiveness on the CSES between the two groups, the Highest Quartile group being more assertive than the Lowest Quartile. In the Lowest Quartile the group categorized as Androgynous on the BSRI would have a higher mean assertiveness score on the CSES than the group in 37 the Lowest Quartile categorized as Feminine and Undifferentiated on the BSRI. Because activity-passivity is not perfectly assessed by the Mf Scale it was expected that there would be only a moderate, though positive, correlation be- tween Mf Scale scores for the entire sample and the scores on the assertiveness measure, CSES. That is, as the Mf scores move in the masculine direction a greater degree of assertiveness, as measured by the CSES, would be evidence. Considering the BSRI category in relation to CSES assertiveness scores, it was hypothesized that the group categorized as Androgynous or Masculine on the BSRI would have a significantly higher mean asser- tiveness score on the CSES than the group categor- ized as Feminine or Undifferentiated. Based on the findings of Black (1956) regarding the adjectives associated with high and low 55' and replicating his individually relative High-Point and Low-Point method, the following predictions were made: A. It was expected that significantly more of the High-Point 5 group would be categorized as Androgynous or Masculine on the BSRI than the Low-Point 5 group, and that significantly more of the Low-Point 5 group would be categorized as 38 Feminine on the BSRI than the High-Point 5 group. B. It was expected that the High-Point 5 group would show a significantly higher mean assertiveness score on the CSES than the Low-Point 5 group. C. The Low-Point 5 group categorized as Androgynous was expected to have a significantly higher mean assertiveness score on the CSES than the Low- Point 5 group categorized as Feminine on the BSRI. Hypotheses 4A through 4C test the same ideas as Hypotheses 1A through 1C using a different method for categorizing the Mf Scale scores as masculine or feminine. Hypotheses Related to Clinical Assumptions Given that the sample was taken from a normal college population the possibility was recognized that the number of profiles meeting the criteria to test the following two by— potheses would not be sufficient for statistical analysis; therefore, these hypotheses were not considered to be an essential part of the study. 5. Dahlstrom, et a1. (1972) states: "The higher the elevation of the neurotic triad, the lower the value of Scale 5." A. Based on the above assumption it was predicted that the 3-point configurations whose three highest scores were on Scales 1, 2, and 3 (i.e., 123, 321, 312, 132, 213), would have a signifi- cantly lower mean score on Scale 5 than the rest 39 of the sample whose highest three points do not include either 1, 2, or 3. The various pattern groupings of 1, 2, and 3 were to be individually compared to the sample mean, given sufficient numbers of subjects per group. A high neurotic triad is believed to be accom- panied by a masochistic trend, the definition of which includes "long-suffering sacrifice" and "unnecessary assumption of burdens" (Dahlstrom, et al., 1972). Therefore, it was predicted that the group whose three highest points included 1, 2, and 3 would have a significantly lower mean assertiveness score on the CSES than the rest of the sample whose highest 3 points did not include 1, 2, or 3. The group with a high 3-point neurotic triad was expected to have significantly more subjects categorized as Feminine on the BSRI than the rest of the sample whose 3 highest points did not in— clude 1, 2, or 3. Of those with a high 3-point neurotic triad (1, 2, 3), the subgroup categorized as Feminine on the BSRI would have a significantly lower mean assertiveness score on the CSES than would the Androgynous, Masculine, or Undifferentiated subgroups. 40 It is further hypothesized that the predictions expressed in Hypotheses 5A through 5D will hold when both relative and absolute elevation are con— sidered. This will be tested by separately analyz- ing profiles which are "clinically significant," i.e., the high neurotic triad group with one or more T scores at or above 70 would be analyzed separately from the neurotic triad group not meeting the criterion for clinical significance. CHAPTER III METHOD Subjects Four hundred and forty-five college women partici- pated by completing the MMPI (Hathaway & McKinley, 1943), the Bem Sex-Role Inventory (Bem, 1974), and the College Self-Expression Scale (Galassi, et al., 1974). Of this number there were 419 correctly completed sets of question- naires used in the data analysis. Participants were obtain- ed from Psychology 160 and 170 classes at Michigan State University during fall, winter and spring terms of 1979-80; they were compensated for their time by receiving extra credit points toward their final grade. Materials and Procedures Participants completed the questionnaires in class- room-size groups. The study was explained in both verbal and written form as "an assessment of personality character- istics of college women to aid in the validation of a fre- unently used psychological test. It will involve filling out three paper-and-pencil personality questionnaires and will take approximately 1 1/2 hours." More specific ques- tions regarding the nature of the study were deferred until all subjects had completed participation. Results in the form of an interpretive summary were made available through 41 42 the Human Subjects Coordinator, Room 135, Snyder Hall. After consent was obtained the participants received a questionnaire packet containing the MMPI, the BSRI, and the CSES. All participants completed the MMPI first. One- half of the participants then completed the BSRI followed by the CSES; the other half then completed the CSES follow- ed by the BSRI. Respondents were identified by an assigned number for purposes of data analysis. Scoring of the Measures MMPI - The MMPI responses were recorded on the standard answer sheet available through the Psychological Corporation. The profiles were scored by hand using tem- plates. Scores on all MMPI scales were obtained for all respondents; for Scale 5 the group mean and approximate quartile cut-off points were calculated and are listed in Appendix A. The Bem Sex-Role Inventory - The BSRI consists of 60 traits, 20 previously rated by college students as feminine, 20 as masculine, and 20 as neutral. Respondents rated them- selves on these traits using a 7-point scale ranging from 1 ("never or almost never true") to 7 ("always or almost always true"). Responses were recorded on Michigan State University data sheets to facilitate computer scoring, as were responses to the CSES. Bem and Watson's revised scoring procedure (Bem & Watson, Note 1, 1976), was applied to the BSRI data. 43 This procedure makes the necessary distinction between an- drogynous and undifferentiated subjects (Spence, et al., 1974, 1975). An individual's Masculinity score is the mean of her self-ratings on the 20 masculine adjectives; the Femininity score is the mean of the 20 feminine adjectives. Using the list of the participants' mean Masculinity and Femininity scores, group medians for Masculinity and Feminin- ity scores were determined. Each person was then classified into a Bem sex role category based on the median Masculinity and Femininity scores of the sample, as shown below: Masculinity Score Above Median Below Median Above . . Median Androgynous Feminine Femininity Score Below Masculine Undifferentiated Median A copy of the BSRI and the scoring key are included in Appendix C. The College Self-Expression Scale - The CSES contains 50 items, 21 of which are positively worded and 29 of which are negatively worded. Participants responded to each item using a 5-point scale: 0, "almost always or always"; 1, "usually"; 2, "sometimes"; 3, "seldom"; 4, "never or rarely." Negatively worded items were reversed scored and the total 44 scores have a theoretical range of 0-250. A low total score is indicative of a generalized nonassertive response pattern. CHAPTER IV RESULTS In order to examine the clinical assumptions regard- ing women's Mf scores on the MMPI, three types of scores were obtained for each participant: (1) MMPI scale scores; (2) a Bem Sex Role Inventory category score; and (3) a College Self-Expression Scale score. Table 1, below, shows the mean scores on the 13 MMPI scales for the group of 419 college women. Table 1. Raw Score Means with K Correction and T Score Equivalents for College Women (N = 419) on the Validity and Clinical Scales of the MMPI Scale Mean T Score of Mean L 2.82 45 F 6.59 59 K 12.98 51 Hs 13.80 51 D 20.56 52 Hy 21.33 54 Pd 22.54 58 Mf 36.82 50 Pa 10.54 58 Pt 29.83 57 Sc 30.07 61 Ma 23.49 66 Si 26.72 52 45 46 For each participant an MMPI profile was plotted and the individual's highest point and two-point codes were determined. The two-point code was based on the relative height of the clinical scale T scores within the individual's profile. Frequencies of the various two-point codes can be found in Appendix B, Table B-1. Also in this appendix (Table B-2) is a listing of the highest single point codes with the frequency and percentage of the profiles that have one or more scores at or above 70 T. Those profiles with at least one score at or above 70 T are considered to be "clinically significant." With the clinically significant designation three ratings are applied: A profile was rated "C" if three or more scores on the clinical scales are at or above 70 T; a "c2" rating was applied if the profile con- tained two clinical scales with scores at or above 70 T; and "cl" was applied if there was only one clinical scale score at or above 70 T. The mean score on the College Self-Expression Scale (CSES) for this sample was 127.94 with a standard deviation of 19.86. The range of scores on this measure was 70 to 182. On the Bem Sex Role Inventory (BSRI) the median Femin- inity score of the sample was 5.10 with a standard deviation of .56; the median Masculinity score was 4.75 with a stand- ard deviation of .76. Table 2 shows the number of partici- pants classified in each sex role category according to the median split method of scoring (Bem & Watson, 1976). 47 Table 2. Number of Participants in Each Category of the Bem Sex Role Inventory Femininity Masculinity Score Score Above 4.75 At or below 4.75 Above 5.10 Androgynous Feminine N = 103 N = 103 At or Below 5.10 Masculine Undifferentiated N = 106 N = 107 The median Masculinity and Femininity scores for the sample are based on a listing of individual's means for the 20 traits comprising the respective scales. Results of Hypothesis Testing Hypotheses 1A through 1C used the quartile method of categorizing Mf scores to look at the Mf Scale in relation to scores on the BSRI and the CSES. Hypothesis 1A - This hypothesis examined the associa- tion between Highest and Lowest Mf Scale quartiles and BSRI categories. A 2 X 2 chi-square analysis was performed to test the hypothesis that there would be a greater frequency of Masculine and Androgynous subjects in the Highest Mf Quartile ("masculine" direction), and a greater frequency of Feminine subjects in the Lowest Mf Quartile ("feminine" direction). Highest and Lowest Quartile groups were based on the Mf scores of the entire sample. Table 3 shows the frequency of subjects in each cate- gory used in the analysis. 48 Table 3. Frequency of Subjects in the Lowest and Highest Mf Quartiles Categorized as Masculine or An- drogynous and as Feminine on the Bem Sex Role Inventory BSRI Category Lowest Mf Quartile Highest Mf Quartile Masculine or Androgynous N = 59 N = 67 Feminine N = 35 N 24 The‘x2 was found to be 2.51, which was not significant at p <<.05 with df = 1. It was thought that the lack of sig- nificance may have been due to the inexact and sometimes arbitrary classification of subjects on the BSRI which results from using the median split technique. Subjects were reclassified using Bem's norms for the Masculinity and Femininity Scales which were obtained in her 1976 Stanford sample. The chi-square for this analysis was larger, 52 = 3.08, df = 1, yet still did not reach significance at p <:.05. Further post-hoc examination of the data revealed the lack of significance as due to similarity in Mf quartile classification for the Feminine and Androgynous subjects. The Feminine and Androgynous groups each had 35 subjects in the Lowest Mf Quartile and 24 subjects in the Highest Mf Quartile. Due to this unexpected lack of difference between the Feminine and Androgynous groups a post-hoc chi-square an- alysis using the Masculine and Feminine groups was performed. For this analysis x? = 6.95, which is significant at p ‘<.01, with df = 1. These results modify the hypothesis, neither 49 directly supporting nor greatly contradicting it. It was expected that, using Mf extreme score groups, BSRI Masculine and Androgynous subjects would score similarly and in the more "masculine" direction on the Mf Scale than the BSRI Feminine group. Instead, it was the Feminine and Androgy- nous subjects who scored similarly on Mf, yet in the less masculine direction on Mf as predicted. Also, it was expected that the range of Mf scores for women categorized as Androgynous on the BSRI would be con— siderable, but not necessarily greater than that of women categorized as Feminine or Masculine. For the Androgynous group the raw Mf scores ranged from 25 to 47; for the Femi- nine group scores ranged from 27 to 49; the Masculine group's range was 26 to 46. (A high raw score is thought to imply high femininity on the Mf Scale.) All BSRI sex role groups showed a wide range of scores on the Mf Scale; however, as the chi-square analysis showed the Masculine group had sig- nificantly more subjects in the Highest, more "masculine," Mf Quartile. The Feminine and Androgynous groups had signif- icantly more subjects in the Lowest, more "feminine," Hf Quartile. Hypothesis 1B - This hypothesis compared the Lowest and Highest Mf Quartile for differences in mean CSES asser- tiveness scores. The clinical assumption that Mf scores in the "feminine" direction are associated with passivity, sub- missiveness, and a masochistic acceptance of discomfort is 50 tested by this comparison. To examine this assumption the mean CSES assertiveness scores for the quartiles were com— pared using a petest. The expectation was that the Lowest Quartile, highly feminine women on the Mf Scale would have significantly lower assertiveness scores than the Highest Quartile, highly masculine women. This hypothesis was not supported as the scores in Table 4 clearly show (p = .10, df = 233). A high score on the CSES is thought to be asso- ciated with a high degree of assertiveness. Table 4. Mean Assertiveness Scores on the College Self- Expression Scale for the Lowest and Highest Mf Quartile Groups Lowest Mf Quartile Highest Mf Quartile Mean CSES Scores 129.13 (N = 120) 128.92 (N = 115) Hypothesis 1C - This hypothesis examined BSRI cate- gory subgroups of the Lowest Mf Quartile for differences in CSES scores. For the Lowest Mf Quartile it was expected that the group categorized as Androgynous on the BSRI would be more assertive on the CSES than the BSRI combined Femi- nine and Undifferentiated group. A p—test performed on the means of the CSES revealed a significant difference between the Androgynous group and the combined Feminine and Undiffer- entiated group in the direction supporting the hypothesis (3 = 3.75, df = 93, p .m mm.m vm. mv. we. we. mH.H em. ..... o Hm.ma mm. mo.v mv.a mm.m mm.a vm.m em. em. ..... m mm.mH mm. mm.aa an. mH.m an. mm. an. we. ..... m mo.oa mm. mm.~ vm.m em. mm.m vm. we. ..... n ow.ma mm. oH.h Hm.a we. mm. mv.a «N. an. em. ..... m ~>.m vm. mm.m an. mm. em. ..... m mn.ma mm. mm.m mH.H em. mm. an. an. mv.H ..... v Ha.m mv. mm.~ mH.H an. em. mv.H we. mm. ..... m oo.v mH.H mH.H me. am. «N. «N. «N. em. ..... m mo.m me. an. an. am. an. «N. ..... a Hobo? o m m h m m w m m H unflom ucflom ocooom 688m 658 83 3.800 350m sown mo Ham comm 53:3 5 a: u E owe—03 omoHHoo 58w moooo mo mmoucmouom .Tm manna. 80 Hoummmnoomaozfiu H .B on mponm no um 800m .H. moo u o as on 909.. o a. or 90%.. no um mum mmuoom .H. mug Ho mg mun U ooumu ma mHCOHm m I. .m. 2. 908... Mo um mum mouoom B mug no woo we ucooflflcmflm handguns oouooflmcoo mwojmoum .4 .1 ma am we om om «H mm mm om Ham v Goa me m we m ma m OH m me o H OH H m o m 0 mm ma H a o o m m a m an H q o a o m mas om mna mm ma mm am we NH om o m m a o m a m N Hanna yahoo ummemflm bemoan. Ema >383an omficoonom .pcmOAwflamAm haemofleflao Hmong kkbommuwu moaaoflmflcmflm Sunflcflao 5:95on a: u E g 68:8 58o 8868va £398 “anagram Eamoflflo on... 9:8 883: .mum 29a. APPENDIX C THE BEM SEX ROLE INVENTORY AND SCORING KEY APPENDIX C THEBEMSEXROIEINVENIOMANDSCORINGIEY AGE CLASS IEVEL (FRESHMAN, SOPHOMDRE, JUNIOR, SENIOR) ACADEMIC MAJOR m the following page, you will be shown a large number of personality characteristics. We would like you to use those character- istics in order to describe yourself. That is, we would like you to indicate, on a scale from 1 to 7, how true of you these various char- acteristics are. Mark your answers with lead pencil on the computer soorable form provided; please do not leave any characteristic unmarked. Example: sly Mark a 1 if it is NEVERORAIJVDSTNFNERTRUE thatyouare sly. Mark a 2 if it is USUALLY NOT TRUE that you are sly. Mark a 3 if it is SOMETD’iES BUT INF‘REQUEN'ILY TRUE that you are sly. Mark a 4 if it is OCEASIONAILY TRUE that you are sly. Mark a 5 if it is OFTEN TRUE that you are sly. Park a 6 if it is LEUALLY TRUE that you are sly. Mark a 7 if it is Ali/MS OR ADDST ALWAYS TRUE that you are sly. Thus if you feel it is sometimes but infrequently true that you are "sly," mark a 3. If it is never or ahrost never true that you are "malicious," mark a 1 and if it is always or almst always true that you are "irresponsible," mark a 7. 81 82 l 2 3 4 5 6 7 NEVER OR USUALLY SOMETIMES OCCASION- OFTEN USUALLY ALWAY S OR AIMDST DUI BUT ALLY TRUE TRUE ALMBT NEVER TRUE TRUE INFREDUEN'ILY TRUE ADAIAY S TRUE TRUE Please mark responses on camputer-soorable fonn. M 1. Self reliant M 22. Analytical F 41. Warm F 2 . Yielding F 23 . Synpathetic N 42 . Solemn N 3. Helpful N 24. Jealous M 43. Willing to M 4. Defends own M 25. Has leadership :akeia beliefs abilities F 5. Cheerful F 26. Sensitive to the F 44' Tender N 6. M 1 needs of others N 45. Friendly M 7. Independent N 27. Truthful M 46. Aggressnre M 28. Willi to take F 47. Gullible F 8' Shy risk:g . . N 9 Conscientious N 48. Ineff1c1ent ' F 29. Understanding M 10 Athletic M 49. Acts as a ' N 30. Secretive leader F 11’ Affecu‘mte 31. Makes decisions F 50. Childlike N 12. Theatrical ea311y N 51. Ada I lle M 13. Assertive F 32. Oanpass1onate M 52. Individual- F 14. Flatterable N 33. Sincere istic N 15. Happy M 34. Self-sufficient F 53. Does not use M 16. Strong personality F 35. Eager to soothe 31:311 e F 17 a1 hurt feelings guag ' 10y . N 54. Unsystematic . N 36. Conceited N 18. Unpredlctable . . . M 55. Oaupet1t1ve M 37. Dominant M 19. Foroeful I F 38 Soft-spoken F 56' S F 20. Feminine ' children N 21. Reliable N 39' ”able N 57. Tactful M 40' Masculine M 58. Ambitious M 59. Gentle N 60. Conventional APPENDIX D ANALYSIS OF VARIANCE SUMMARY TABLES APPENDIX D ANALYSIS OF VARIANCE SUMMARY TABLES Table D-l. Analysis of Variance Relating Bem Scales to MMPI High and Low Point Mf Groups Source SS df MS F Mf Group (A) .35 1 .35 1.59 Bem Scale (B) 2.09 l 2.09 9.5* AB 2.05 1 2.05 9.3* S/AB 114.05 524 .22 *p '<.01 Table D-2. Analysis of Variance Relating Bem Sex Role Inventory Categories to College Self-Expression Scores Source SS df MS F Bem Cate- gory UU 27603.86 3 9201.29 34.29* S/A 11353.09 415 268.32 * p < .001 83 APPENDIX E CSES AND BSRI STATISTICS FOR THE MF SCALE GROUPS APPENDIX E CSES AND BSRI STATISTICS FOR THE MF SCALE GROUPS Table E-l. Lowest Mf Quartile Statistics (Feminine) Mf Scale CSES BSRI Mean = 42.03 Mean = 129.13 Category Frequency Standard Standard Masculine Deviation = 1.94 Deviation = 12.98 24 Raw Score Androgynous Range = 40 - 49 35 N = 120 Feminine 35 Undifferentiated 26 Table E-2. Highest Mf Quartile Statistics (Masculine) Mf Scale CSES BSRI Mean = 31.37 Mean = 128.92 Category Frequency Standard Deviation = 2.61 Standard Masculine Deviation = 19.37 43 Raw Score Androgynous Range = 21 - 34 24 N = 115 Feminine 24 Undifferentiated 24 84 85 Table E-3. Lowest Point Mf Group Statistics (Feminine) Mf Scale CSES BSRI Mean = 40.95 Mean = 123.53 Category Frequency Standard Standard Masculine Deviation = 3.05 Deviation = 19.00 24 Raw Score Androgynous Range = 34 - 49 30 N = 132 Feminine 40 Undifferentiated 38 Table E-4. Highest Point Mf Group Statistics (Masculine) Mf Scale CSES BSRI Mean = 29.60 Mean = 133.15 Category Frequency Standard Standard Masculine Deviation = 3.04 Deviation = 19.43 8 Raw Score Androgynous Range = 21 - 36 4 N = 20 Feminine 3 Undifferentiated 5 APPENDIX F THE COLLEGE SELF-EXPRESSION SCALE APPENDIX F THE COLLEEE SHE-MESS ION SCALE The following inventory is designed to provide information about the way in which you express yourself. Please answer the questions by checking the appropriate box from 0—4 (Almost Always or Always, 0; Usually, 1; Sorretjmes, 2; Seldan, 3; Never or Rarely, 4) on the carputer answer sheet. Your answer should reflect hon you generally express yourself in the situation. 10. ll. 12. Do you ignore it when someone pushes in front of you in line? When you decide that you no longer wish to date saneone, do you have marked difficulty telling the person of your decision? Would you exchange a purchase you discover to be faulty? If you decided to change your major to a field which your parents will not approve, would you have difficulty telling them? Are you inclined to be over-apologetic? If you were studying and if your roommate were making too much noise, would you ask him to stop? Is it difficult for you to compliment and praise others? If you are angry at your parents, can you tell them? Do you insist that your roonmate does her fair share of the cleaning? If you find yourself becoming fond of scmeone you are dating, would you have difficulty expressing these feelings to that person? If a friend who has borrowed $5.00 from you seems to have for- gotten about it, would you remind this person? Are you overly careful to avoid hurting other people's feelings? 0 1 2 3 4 Almost al- Usually Sane— Seldan Never or ways or tines rarely alwa s y 86 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 87 If you have a close friend whom your parents dislike and con- stantly criticize would you inform your parents that you disagree with them and tell them of your friend's assets? Do you find it difficult to ask a friend to do a favor for you? If food which is not to your satisfaction is served in a restau- rant, would you conplain about it to the waiter? If your roomrete without your permission eats food that he knows you have been saving, can you express your displeasure to him? If a salesman has gone to considerable trouble to show you sone merchandise which is not quite suitable, do you have difficulty in saying no? Do you keep your opinions to yourself? If friends visit when you want to study, do you ask them to return at a more convenient tine? Are you able to express love and affection to people for whom you care? If you were in a small seminar and the professor made a staterrent that you considered untrue, would you question it? If a person of the Opposite sex whom you have been wanting to neet smiles or directs attention to you at a party, would you take the initiative in beginning a conversation? If someone you respect expresses opinions with which you strongly disagree, would you venture to state your own point of View? Do you go out of your way to avoid trouble with other people? If a friend is wearing a new outfit which you like, do you tell that person so? If after leaving a store you realize that you have been "short- changed," do you go back and request the correct arrount? If a friend nekes what you consider to be an unreasonable request, are you able to refuse? 0 l 2 3 4 Alnost al- Usually Sorre- Selda'n Never or ways or tines rarely always 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 88 If a close and respected relative were annoying you, would you hide your feelings rather than express your annoyance? Ifyourparentswantyoutocomehome foraweekendbutyouhave made important plans, would you tell them of your preference? Do you express anger or annoyance toward the opposite sex when it is justified? If a friend does an errand for you, do you tell that person how much you appreciate it? men a person is blatantly unfair, do you fail to say sanething about it to him/her? Do you avoid social contacts for fear of doing or saying the wrong thing? If a friend betrays your confidence, would you hesitate to express annoyance to that person? menaclerkinastorewaitsonsomeonewhohasccmeinafter you, do you call his attention to the matter? If you are particularly happy about someone's good fortune, can you express this to that person? Would you be hesitant about asking a good friend to lend you a few dollars? If a person teases you to the point that it is no longer fun, do you have difficulty expressing your displeasure? If you arrive late for a meeting, would you rather stand than go to a front seat which could only be secured with a fair degree of conspicuousness? If your date calls on Saturday night 15 minutes before you are supposed to meet and says that he has to study for an important exam and cannot make it, would you express your annoyance? If someone keeps kicking the back of your chair in a movie, would you ask him/her to stop? If someone interrupts you in the middle of an important conversa- tion, do you request that the person wait until you have finished? 0 1 2 3 4 Almost al- Usually Some- Seldom Never or ways or times rarely always 43. 44. 45. 46. 47. 48. 49. 50. 89 Do you freely volunteer information or opinions in class discussions? Are you reluctant to speak to an attractive acquaintance of the Opposite sex? If you lived in an apartment and the landlord failed to make certain necessary repairs after promising to do so, would you insist on it? Ifyourparentswantyouhorebyacertaintimewhichyoufeel is much too early and unreasonable, do you attempt to discuss or negotiate this with them? Do you find it difficult to stand up for your rights? If a friend unjustifiably criticizes you, do you express your resentment there and then? Do you express your feelings to others? Do you avoid asking questions in class for fear of feeling self- conscious? 0 1 2 3 4 Almost al- Usually Sore- Seldom Never or ways or times rarely always REFERENCES REFERENCES Alberti, R. E., & Emmons, M. L. Your perfect right: A gpide to assertive behavior. San Luis Obispo, CA: Impact, 1970. Barger, B., & Hall, E. Personality patterns and achieve- ment in college. Educational and Psychological Bechtoldt, H. P., & Dahlstrom, W. G. MMPI scale intercor- relations and high-point codes on college freshmen. Unpublished materials, 1953. Cited in W. G. Dahl- strom, G. S. Welsh, & L. E. Dahlstrom, An MMPI handbook (Vol. 1). Minneapolis: Univers1ty of Minnesota Press, 1972. Bem, S. L. The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 1974, 13! 155-162. Bem, S. L. Sex role adaptability: One consequence of psy- chological androgyny. Journal of Personality and Social Psychology, 1975, ll, 634-643. Bem, S. L. On the utility of alternative procedures of assessing psychological androgyny. Journal of Con- sulting and Clinical Psychology, 1977, 55, 196-205. Bem, S. L., Martyna, W., & Watson, C. Sex typing and an- drogyny: Further explorations of the expressive domain. Journal of Personality and Social Psy- chology, 1976, 35! 1016-1023. Berzins, J., Welling, M., & Wetter, R. A new measure of psychological androgyny based on the Personality Research Form. Journal of Consulting and Clinical Psychology, 1978, 49, 126-138. Biller, H. B. Paternal and sex-role factors in cognitive and academic functioning. In J. K. Cole & R. Dienstbier (Eds.), Nebraska Symposium on Motivation (Vol. 21). Lincoln: University of Nebraska Press, 1973. Black, J. D. Adjectives associated with various MMPI codes. In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic read- ings on the MMPI in psychology and medicine. Minneapolis: University of Minnesota Press, 1956.(a) 90 91 Black, J. D. MMPI results for fifteen groups of female college students. In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI in psychology and medicine. Minneapolis: University of Minnesota Press, 1956. (b) Bonder, G. E. The role of assessment in assertion train- ing. Counseling Psyohologist, 1975, 5, 90-96. Brehem, M. L. An examination of MMPI results and delin- quency data for college and non-college groups. Unpublished manuscript, 1954. Cited in W. G. Dahl- strom, G. S. Welsh, & L. E. Dahlstrom, An MMPI hand- book (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Broverman, I. K., Vogel, S. R., Broverman, D. M., Clarkson, F. E., & Rosenkrantz, P. 8. Sex role stereotypes: A current appraisal. Journal of Social Issues, 1972, 38 (2), 59-78. Burkhart, B. R., Green, S. B., & Harrison, W. H. Measure- ment of assertive behavior: Construct and predic- tive validity of self-report, role playing, and in- vivo measures. Journal of Clinical Psychology: 1979, 35, 376-383. Carson, R. C. Interpretive manual to the MMPI. In J. N. Butcher (Ed.), MMPI: Research, development, and clinical applications. New York: McGraw-Hill, 1969. Consentino, R., & Helibrun, A. B., Jr. Anxiety correlates of sex-role identity in college students. Psycho- logical Raports, 1964, 121 729-730. Constantinople, A. Masculinity-femininity: An exception to a famous dictum. Psychological Bulletin, 1973, 89, 389-407. Dahlstrom, W. G., & Reipler, C. B. MMPI code patterns for entering freshmen. Unpublished materials, 1970. Dahlstrom, W. G., Welsh, G. S., & Dahlstrom, L. E. Ag MMPI handbook (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Drake, L. E. Differential sex responses to items of the MMPI. Journal of Applied Psychology: 1953, 21, 46. 92 Eisler, R. M., Miller, P. M., & Hersen, M. Components of assertive behavior. Journal of Clinical Psychology, 1973, E2, 295-299. Eysenck, H. J., & Eysenck, S. B. Manual fopthe Eysenck Personality Inventopy. San Diego: Educational and Industrial Testing Service, 1968. Fry, F. D. A study of the personality traits of college students and of state prison inmates as measured by the MMPI. Journal of Psychology, 1949, $8, 439-449. Galassi, J. P., DeLo, J. S., Galassi, M. D., & Bastien, S. The College Self-Expression Scale. Behavior Therapy, 1974, 5, 165-171. Galassi, J. P., & Galassi, M. D. Alienation in college students: A comparison of counseling seekers and nonseekers. Journal of Counseling Psychology, 1973, 20, 44-49. Galassi, J. P., & Galassi, M. D. Validity of a measure of assertiveness. Journal of Counseling Psyohology, Gaudreau, P. Factor analysis of the Bem Sex-Role Inventory. Journal of Consulting and Clinical Psychology, 1977, 35, 299-302. Gayton, W. F., Havu, G., Ozmon, K. L., & Tavormina, J. A comparison of the Bem Sex-Role Inventory and the PRF ANDRO Scale. Journal of Personality Assessment, 1977, 211 619-621. Gough, H. G. Identifying psychological femininity. Educa- tional and Psycholpgical Measurement, 1952, 12, 427-439. — Gough, H. G., & Heilbrun, A. B. The Adjective Check List manual. Palo Alto, CA: Consulting Psychologists Press, 1965. Graham, J. R., Schroeder, E. E., & Lilly, R. S. Factor analysis of items on the Social Introversion and Masculinity-Femininity scales of the MMPI. gggrnal of Clinical Psychology, 1971, 31, 367-370. Haglund, S. V. Relationships among adjustment variables and sex role orientation in college women: A construct validation study of psychological androgyny. Disser- tation Abstracts International, 1978, 38 (8-B), 3880- 3881. 93 Harford, T. C., Willis, C. H., & Deabler, H. L. Personality correlates of masculinity-femininity. Psychological Reports, 1967, ll, 881-884. Hathaway, S. R. Scales 5 (masculinity-femininity), 6 (para- noia), and 8 (schizophrenia). In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI inlpsy- chology and medicine. Minneapolis: University of Minnesota Press, 1956. Hathaway, S. R., & McKinley, J. C. A multiphasic personal- ity schedule (Minnesota): I. Construction of the schedule. Journal of PsychOlogy, 1940, lg, 249-254. Hathaway, W. R., & McKinley, J. C. The Minnesota Multi- phasic Personalitlenventory. New York: Psycholog— ical Corporation, 1943. Hathaway, 8. R., & Meehl, P. E. Adjective check list cor- relates of MMPI scores. Unpublished materials, 1952. Cited in W. G. Dahlstrom, G. S. Welsh, & L. E. Dahl- strom, An MMPI handbook (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Heilbrun, A. B., Jr. Personality differences between ad- justed and maladjusted college students. Journal of Applied Psychology: 1960, 45, 341-346. Heilbrun, A. B., Jr. Parent identification and filial sex- role behavior: The importance of biological context. In J. K. Cole & R. Dienastbier (Eds.), Nebraska Symposium on Motivation (Vol. 21). Lincoln: University Press, 1973. (Heilbrun, A. B., Jr. Measurement of masculine and feminine sex role identities as independent dimensions. Journal of Consulting and Clinical Psyghology, 1976, 55! 183-190. Hogan, H. W. The measurement of psychological androgyny: An extended replication. Journal of Clinical Psy- chology, 1977!.32! 1009-1013. Jackson, D. N. Personality Research Form manual. Goshen, NY: Research Psychologists Press, 1967. Kelly, J. A., & Worell, J. New formulations of sex roles and androgyny: A critical review. Journal of Con- sulting and Clinical Psychology, 1977, 23! 1101-1115. Keppel, G. Design and analysis--a researchers' handbook. Englewood Cliffs, NJ.: PrentiCe-Hall, 1973. 94 Kipper, D. A., & Jaffee, Y. Dimensions of assertiveness; Factors underlying the College Self-Expression Scale. Perceptual and Motor Skills, 1978, fig, 47-52. Klopfer, W. G. Correlations of women's Mf scores on the MMPI and Strong VIB. Journal of Clinical Psychology, 1966, go, 216. Kokosh, J. MMPI personality characteristics of physical and social science students. Psychological Reports, 1969, 31! 882-893. LaChar, D. The MMPI: Clinical assessment and automated interpretation. Los Angeles: Western Psychological Services, 1974. Lanyon, R. I. Psychological Screening Inventory manual. New York: Research Psychologists Press, 1973. LaTorre, R. A. The psychological assessment of gender identity and gender role in schizophrenia. Schizo- phrenia Bulletin, 1976, g, 266-285. LaTorre, R. A. Gender role and psychological adjustment. Archives of Sexual Behavior, 1978, 1, 89-96. LaTorre, R. A., & Gregoire, P. A. Gender role in university mental health clients. Journal of Individual Psy- chology, 1977, ll, 246-249. Little, J. W. An analysis of the MMPI. Unpublished thesis. University of North Carolina, 1949. Cited in W. G. Dahlstrom, G. 8. Welsh, & L. E. Dahlstrom, An MMPI handbook (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Mills, W. W. MMPI profile pattern and scale stability throughout four years of college attendance. Un- published doctoral dissertation, University of Minnesota, 1954. Cited in W. G. Dahlstrom, G. W. Welsh, & L. E. Dahlstrom, An MMPI handbook, (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Mischel, W., & Mischel, H. The nature and development of psychological sex difference. In G. S. Lesser (Ed.), Psychology and educationallpractice. -Glenview, IL: Scott Foresman, 1971. Murray, J. B. The Mf scale of the MMPI for college students. Journal of Clinical Psychology, 1963, l9, 113-115. 95 Mussen, P. H. Long-term consequents of masculinity of interests in adolescence. Journal of Consulting Psychology: 1962, 39, 435-440. Nevill, D. D. Sex roles and personality correlates. Human Relations, 1977, 29' 751-759. O'Leary, K. D., & Wilson, G. T. Behavior therapy: Appli- cation and outcome. Englewood Cliffs, NJ: Prentice- Hall, 1975. Pepper, L. J., & Strong, P. N. Judgmental subscales for the Mf scale of the MMPI. Unpublished materials, 1958. Cited in W. G. Dahlstrom, G. S. Welsh, & L. E. Dahlstrom, An MMPI handbook (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Rathus, S. Principles and practices of assertive training: An eclectic overview. Counseling Psychologist, 1975, 5, 9-20. ‘ Rosenkrantz, P. S., Vogel, S. R., Bee, H., Broverman, I.K., & Broverman, D. M. Sex-role stereotypes and self- concepts in college students. Journal of Consulting and Clinical Psychology, 1968, ll, 287-295. Schiff, E. The relationship of women's sex-role identity to self-esteem and ego development. Dissertation Abstracts International, 1977, l8 (6-B), 2838. Schofield, W. Changes in responses to the MMPI following certain therapies. Psychologlcal Monographs, 1950, pg, (5, Whole No. 311). Schofield, W. A further study of the effects of therapies on MMPI responses. Journal of Abnormal and Social PsycholOQY: 1953, $8, 67-77. Sears, R. R. Relation of early socialization experiences to self-concepts and gender role in middle child- hood. Child Development, 1970, ll, 267-289. Spence, J. T., Helmreich, R., & Stapp: J. The Personal Attributes Questionnaire: A measure of sex role stereotypes and masculinity-femininity. JSAS Cata- logue of Selected Documents in Psychology, 1974, g, 43. Spence, J. T., Helmreich, R., & Stapp, J. Ratings of self and peers on sex role attributes and their relation- ship to self-esteem and conceptions of masculinity and femininity. Journal of Personality and Social Psychology, 1975, 33, 29-39. 96 Stebbins, C. A., Kelly, B. R., Tolor, A., & Power, M. Sex differences in assertiveness in college students. Journal of Psychology, 1977, 95, 309-315. Strahan, F. Remarks on Bem's measurement of psychological androgyny: Alternatives, methods, and a supple- mentary analysis. Journal of Consulting and Clinical Psychology, 1975, 5}, 568-571. Terman, L. M., & Miles, C. C. Sex and personality: Studies in masculinity and femininity. New York: McGraw-Hill, 1936. Terman, L. M., & Miles, C. C. Manual of information and directions for use of the Attitude-Interest Analysis Test. New York: McGraw-Hill, 1938. Tolor, A., Kelly, B. R., & Stebbins, C. A. Assertiveness, sex-role stereotyping, and self-concept. Journal of Psychology, 1976, 22, 157-164. Waters, C. W., Waters, L. K., & Pincus, S. Factor analysis of masculine and feminine sex-typed items from the Bem Sex-Role Inventory. Psychological Reports, 1977, 39, 567-570. Webb, A. P. Sex-role preferences and adjustment in early adolescents. Child Development, 1963, 35, 609-618. Webster, H. Personality development during the college years: Some quantitative results. Journal of Social Issues, 1956, ll, 29-43. Whetton, C., & Swindells, T. A factor analysis of items of the Bem Sex-Role Inventory. Journal of Clinical Psychology, 1977, 2;, 150-153. 97 Reference Note Bem, S. L., & Watson, C. Scoring Packet: Bem Sex- Role Inventory. Revised 4/76. (Available from Sandra L. Bem, Department of Psychology, Stanford University, Stanford, CA, 94305.) "Ill?11114117114471is