.‘ «.4: . 'll‘i.§ tu- vvvs'v :..v.A. 1.7. WV . - .~‘1.J .v.» ..-- x‘_.,. ~r¢ 'W-Z‘x .3. fig: ‘2; ' «5“ ‘ § "— xrx. . a . ‘3: . .p. “'1'.“ C .‘ - E .C’t-lmfxb ‘ , A}? >33 .x 45‘ . ’ 4.x, . 5*: ... .~ ' yin" -. P”) (1).! -\\a‘. . _ ‘ ‘L‘; {‘5‘ “3.; ,5- 7),! ‘ "a" - , #I-‘Ew.’ 7 Mai}: (“W‘ 7 - a": ‘J '>< l' 42,. m. q». 1" +1;— .;:u ‘2‘.“ 42-” 4:.) ’w hc~ . 0' ‘9 r‘ ’ . "I '~h'-',' I '(J "‘r'f’ ”£13,“ I: ’ - ‘10:. fl? .«7. 3 WM; 1mg 1') 1;}C’I i I I}! .. ) ; 2' 3.; C 0"" UN» p5}. :2}: fig; 23' 9;???3'11 'L ., 'rr 5??/J . I I: 5.26:? ’ " I I . - - 3.;- . l' "r," '{1‘ "A"? gingfifiw 43:2: (:14. -. . l n.’ 4.1”. :35r/U’J él‘.{;/ 1!]! 5;— .'.'- . . 1-65 , Mai-a: "$533111“: 4: 1m. ' Mix-W ‘3 ~- < 5:”: .'. V ."'\ “1-11 t ‘V‘ :3... .‘u .3 21- "1.:- fl 3:1;3- -.— “FER. g)” 1-3;» fiS’v‘gfir‘ “‘ .. :‘v. I K. J )‘k 7': .gf- Juav; -;::::“:‘ I‘M-- ' 4 [I 't Li‘r’firéraan"&"’ofil . 5353-4 ’r v.1”? 4;) « . ,3»; :&{-, J: n, h, 1"“..- 3"," r1. ’II;.7 -~'~w"a{f{;m '5 I j w gay/2, ‘1 4:2,; W36" - , ,. , -? ;}:p'.j£! ' ’/ J'é 5%:52? ‘._!‘v /..:‘:}:fgfioln,£lg;i;:¥ fr?” .J. {’Yfic’ .q”:"f7é’?{;f’ '42,.” av:- 42:: - Eifffifli ué‘gdé’éfigalgj/gfigf £37; 2:04 35" at} w‘ n .{m/gjfiy mm a A; .97] r 4.: - Iv? 0".) ‘ ’6 “.: iffi. C;;:v;:. :-:{\;.I Sing: “3:“ k .3". 1.5531. NR. . '7.“ :-'.‘~‘.',{ .~. -. . 48;.“ “‘3‘ ;‘-\n~u.:.g, . "v ' I - n‘ - ~ “'1‘ ~ .. H "1.: 1:...“\r: '13:: - - «“4 L9}; " \'_ oi 2-15“; 2. n ’ 5325:313356." ,,:¢._.‘f‘~ u" h I“. “ .. 2; ,. ’75? ’l I I ) L p-’./ F5 ~t l..]_ 0v a.!? {1.1.3:}: { 1’1”“: {Jé’fia ‘ find if} a! 7." 5:; of/{' gyfr'l: , a. 4 ‘ I I" , 1' ’3'"! . 545?: .I I \32‘ «9,33 “N "121:" 5“, 4,54%? -"’- ’ 'Jié‘g'éfy”: 5,} 25?}. - I J P I: «' V ‘ t :4 v :1 .. :2 5:195- ! ‘31» ‘- ZOSSMI‘7 MICHI IIIIIIIIIIIIIIIIIIIIIIIIIIII 93 00533 0414 LIBRARY Michigan State University This is to certify that the dissertation entitled Jungian Typology in Relation To Gender And Psychopathology: An Exploratory Study presented by Daniel G. Mosher has been accepted towards fulfillment of the requirements for Ph.D. degreein PsychologL )ZQM flit Major profedor Date Q,/[VC8'7 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 I MSU RETURNING MATERIALS: Place in book drop to remove this checkout from LIBRARIES n your record. FINES will be charged if book is returned after the date stamped below. .. -- w” ' Exfifi i? ‘ f.. JUNGIAN TYPOLOGY IN RELATION TO GENDER AND PSYCHOPATHOLOGY: AN EXPLORATORY STUDY BY Daniel Gene Mosher A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1987 Y7 ‘5’/ I" .1 ' _! )p: .I' A, // ABSTRACT JUNGIAN TYPOLOGY IN RELATION TO GENDER AND PSYCHOPATHOLOGY: AN EXPLORATORY STUDY BY Daniel Gene Kosher The purpose of the present study was to explore the potential contribution of Jungian typology and the contrasexual archetypes (anima-animus) to an under- standing of gender differences in psychopathology. The Singer-Loomis Inventory of Personality (SLIP), Symptom Checklist 90R (SCL-90-R), Minnesota Multiphasic Person- ality Inventory (MMPI) Psychopathic Deviate Scale, Impulsiveness Scale, Bem Sex Role Inventory (BSRI), Personal Attributes Questionnaire (PAQ), and the Anima- Animus Scale were administered in counterbalanced order to 120 participants from introductory psychology courses. As predicted, females had significantly higher scores than males for introverted feeling and intro- verted intuition, which in turn, were positively linked with most measures of psychopathology: particularly symptoms of neuroses and depression. Males, however, did not have higher scores than females for extraverted Daniel Gene Mosher thinking and extraverted sensation as expected, nor were the two function types related to symptoms of impulsiveness and sociopathy as predicted. 0n the other hand, females did tend to have somewhat higher scores than males for symptoms of neuroses and depres- sion, whereas males tended to have higher scores for impulsiveness and sociopathy than females. As pre- dicted by Jungian theory, female Animus scores and male Anima scores were negatively associated with psy- chopathology. A comparison with the BSRI and PAQ pro- vided support for the construct validity of the Anima- Animus Scale, but also indicated that the scale mea- sures somewhat distinct aspects of masculinity-feminin- ity. Using a multivariate regression analysis, androg- yny (as measured by the BSRI and FAQ) did not signifi- cantly contribute to a predictive system for psychopathology, over and above the main effects for masculinity. However, the Anima subscale and intro- verted feeling and introverted intuition did signifi- cantly contribute to the predictive system beyond the effects of masculinity. Thus, the present study offers corroboration for the hypothesis that gender differences in typology underlie the documented gender differences in psy- chopathology. Furthermore, Jung's theory of the Daniel Gene Mosher contrasexual archetypes and their relation to psychological health also appears corroborated. The present study, however, offers little support for either an additive or balance theory of androgyny in relation to psychological adjustment. To Diann, with love. ii ACKNOWLEDGMENTS There are many people to whom I am very grateful for making this research project possible. I would like to give special thanks to my chairperson, Dr. Joseph Reyher, who has been so helpful throughout this enterprise. He has shared my excitement for research and truly served as a model of the researcher-clini- cian. Dr. Norman Abeles, who has been very helpful in this work, and also by providing suggestions and guid- ance throughout my graduate career. Dr. Imogen Bowers, who has played a special role in my development as a psychologist, and was the first to recognize there might be a therapist "in there some- where". I also wish to thank Dr. Elaine Donelson, who has shared my interest in a subject area that is special to both of us. Others I wish to thank include: Ginny Hanson for her invaluable help with data collection, Ann-Marie Scheerbaum for her encouragement and understanding; Dr. Robert Caldwell, who has played a special part in my iii graduate training; Dr. William Mueller for the major role he has played in my becoming a therapist; and Cindy Jennings for the excellent job of preparing this manuscript. And most of all, I wish to thank my wife, Diann, who has helped provide the balance and purpose in my life that was essential for me to complete this task. Without her love and sense of humor this would not have been possible. iv TABLE OF CONTENTS LIST OF TABLES. . . . . . . . . Introduction . . . . . . . . . Methods . . . . . . . . . . . Participants . . . Materials . . . . Design . . . Setting and Procedure Results . . . . . . . . . . . Manipulated Effects . . . . . . Correlational Findings . . . . . Multivariate Regression Analysis. . Other Findings. . . Comparison of Masculinity-Femininity Discussion . . . . . . . . . . Measures Masculinity, Femininity, and Androgyny. . . Nature vs. Nurture and Jung . . . Implications . . . . . . . . References . . . . . . . . . . Appendix A: Anima-Animus Scale . . . . . Page vi 13 13 13 19 20 23 29 .33 4O 43 44 50 54 60 68 7O 78 Table 10 LIST OF TABLES Page Means and Standard Deviations (Present and Previous Study) for SLIP Subscale Scores. . . . . . . . . . . . 24 Means and Standard Deviations (Present Study and Norms) for Psychopathology Scales. . . . . . . . . . . . 25 Means and Standard Deviations (Present Study and Norms) for Masculinity- Femininity Scales . . . . . . . . 26 Comparison of Mean Function Type and Mean Psychopathology Scores Between Males and Females (df = 118). . . . . . . . Pearson Product Moment Correlational Analysis of Function Type and Psycho- pathology. . . . . . . . . . . Comparison of Introverted Thinking/ Introverted Sensation and Introverted Feeling/Introverted Intuition Matched Pairs of Items . . . . . . . . . Pearson Product Moment Correlational Analysis of Anima-Animus Scale and Psychopathology. . . . . . . . . Regressions of Masculinity, Femininity, Androgyny, Gender, Function Type, and Pairwise Interactions on SCL-90-R Global Severity Index (GSI) . . . . . . . Pearson Product Moment Correlational Analysis of Anima-Animus, BSRI, and FAQ scales 0 O O O O O O O O O O 0 Pearson Product Moment Correlational Analysis of Psychopathology, Masculinity, Femininity, and Androgyny . . . . . vi 3O 35 37 39 42 46 48 Introduction Two influential studies during the 1970's (Chesler, 1972: Gove & Tudor, 1973) concluded that rates of mental illness are higher for women than for men. While not all subsequent studies (Goldman & Ravid, 1980) have found large sex differences in over- all psychological disorders, the consistency of sex differences in type of psychopathology is striking. There is a clear pattern across studies whereby higher rates of neuroses are reported for women and higher rates of antisocial and addictive types of personality disorders are reported for men (Dohrenwend et a1., 1980). Furthermore, women are more apt than men to claim psychological distress (Brown & Donderi, 1985) and to seek out professional help (Rohrbaugh, 1979). They are twice as likely to develop a severe depression (Amneson & Lewinsohn, 1981; Klerman & Weisman, 1980), and they attempt suicide three times as often (Waldron, 1982). There have been two major approaches to explaining these sex differences in psychopathology (Al-Issa, 1982). One theory (not supported by data) proposes that male and female rates of mental illness are compa- rable, but because of their gender role, females report more illness and are more comfortable with the sick role than males. The second (more data supported) approach suggests that sex differences are caused by stresses related to gender role. It is assumed that because the female role is more stressful than the male role, women tend to have more mental illness than men. Kaplan (1983) argued that an additional explana- tion for these sex differences is that masculine-biased assumptions about what behaviors are healthy and what behaviors are disturbed are codified in diagnostic cri- teria, namely DSM III (American Psychiatric Associa- tion, 1980). These criteria then influence diagnosis and treatment rates and patterns. Williams and Spitzer (1983) pointed out that although Kaplan's view raised important issues, she unfortunately presented no data to support her thesis. The authors examined the sex ratios for DSM III Axis I clinical symptoms and found many categories that were more commonly diagnosed in men, thus providing no support for Kaplan's (1983) the- sis. Kass, Spitzer, and Williams (1983) further argued against Kaplan's view by citing evidence of higher rates of personality disorders in men. In his review, Al-Issa (1982) concluded that the consistently higher rates of neuroses among women and higher rates of personality disorders among men would suggest that sex differences are in the style of reac- tion to, rather than the amount of, stress due to sex role differences. Jung's (1971) theory of psychological types offers a plausible theoretical model for differences in cogni- tive style that may underlie this observed pattern of gender differences in psychopathology. The theory incorporates two bipolar attitudes (introversion and extraversion) which, as the following discussion will show, have been linked to both differences in psy- chopathology and gender. The extraverted attitude is an outward focusing of attention, or libido (psychic energy), toward objects, people, and the environment as the basis of conscious- ness. What the extravert experiences at any given moment is felt to be more directed by the sights or sounds of the objects themselves. The introverted attitude, on the other hand, is an inward turning of the libido toward the subject as the basis of con- sciousness. Introverts attend more to ideas, concepts, impressions, feelings, and images. For the introvert, experience always appears mediated by his or her expe- riencing of it. The distinction between extraverts and introverts must be viewed from the perspective of dif- fering perceptions or ways of processing information. In contrasting introversion and extraversion, Jung wrote: The first attitude (introversion) if normal, is revealed by a hesitating, reflective, ret- icent disposition, that does not easily give itself away, that shrinks from objects, always assuming the defensive, and preferring to make its cautious observations as from a hiding place. The second type (extraversion) if normal, is characterized by an accommodat- ing, and apparently open and ready disposi- tion, at ease in any given situation. This type forms attachments quickly, and ventures, unconcerned and confident, into unknown situ- ations, rejecting thoughts of possible con- tingencies. In the former case, manifestly the subject, in the latter the object, is the decisive factor (p. 545). Elsewhere, Jung (1971) further developed this con- ception: A tense attitude is in general characteristic of the introvert, while a relaxed, easy atti- tude distinguishes the extravert (p. 287). . . To sum up: the primary [most developed] function is in my view more important than the secondary. The intensity of the primary function is the decisive factor. It depends on the general psychic tension, i.e., on the amount of accumulated, disposable libido. The factors determining this accumulation are the complex resultants of all the antecedent psychic states--mode, attention, affect, expectancy, etc. Introversion is character- ized by general tension, an intense primary function [and thus more accumulated or damned up libido]. . .extraversion [is character- ized] by general relaxation, and a weak pri- mary function. . . (p. 288). Psychopathology is produced when the excessive conscious development of one attitude (introversion versus extraversion) blocks the unconscious compen- satory effects of its bipolar opposite in the shadow (the rejected aspects of the personality) where it is subject to repression and regression. Eventually the ego comes under the sway of the unconscious attitude, and when it does, there are emotional and behavioral disturbances. These are more severe and variegated when introversion is the conscious attitude (Mullahy, 1948; Singer, 1973; Whitmont, 1969). According to Jung (1964), femininity (Eros) is linked to introversion in that it fosters introspection and subjectivity. Masculinity (Logos), on the other hand, appears more closely related to extraversion, in that it signifies objective interest, structure, form and discrimination. Thus, consistent with Jung's description of intro- version, females, who tend towards introversion, would be expected to display different patterns of psy- chopathology than males, who tend toward extraversion. Although Jung did not specifically relate extraversion to psychopathology, his description of the extravert as easygoing and failing to think through contingencies, could be extrapolated as a tendency toward impulsivity and acting out behaviors. Eysenck (1960), expanding on the relationship between attitude and psychopathology and using somewhat different definitions of introversion-extraversion, found introversion associ- ated with anxiety, obsessional, and depressive symp- toms, whereas extraversion was associated with psycho- pathic symptoms. In addition, Richek (1969), using the Myers-Briggs Type Indicator (MBTI) (Meyers, 1962), found that extraverts (both moderate and extreme) scored higher on measures of psychological adjustment than introverts, and concluded that extraverts were the least neurotic, and introverts, the most neurotic. Therefore, both studies found relationships between attitude and type of psychopathology, but neither study included gender as a factor. With respect to gender, attitude, and type of psy- chopathology, women, because of their introversion, accumulate disposable (damned up) libido (psychic energy) which is converted to depression and neurotic disorders, including those hysterical disturbances of anorexia, bulimia, dissociation, and multiple personal- ity. Indeed, according to DSM-III (American Psychi- atric Association, 1980), women are assigned these diagnoses more often than men. Conversely, males, because of their extraversion, ought to be more dis- posed to impulsivity and sociopathy (American Psychi- atric Association, 1980), and they are. Research on expectancies and causal attribution is consistent with a correspondence between attitude (introversion-extraversion) and gender. Frieze (1975), has shown that women often expect to fail and tend to internalize the cause of failure. Men generally attribute their success to factors such as ability and their failure to luck (externalize), whereas women tend to attribute failure to lack of ability (internalize) and success to luck or circumstance. Other aspects of Jung's (1964) perspective further elucidate the contribution of gender to type of psy- chopathology. The psychic structure of the two sexes corresponds to a different biological makeup: a pre- dominantly feminine conscious personality in women, masculine in men, together with a predominantly mascu- line (animus) or feminine (anima) component, respec- tively, in the unconscious. He used masculine and fem- inine not to describe roles or stereotypes but, rather, to denote archetypal (instinctual) principles: the feminine principle, eros, the masculine, logos. According to Young-Eisendrath (1985), the contra- sexual elements (animus/anima) constitute an archetypal-complex that operates in the shadow outside personal consciousness. Well-being is dependent upon an integration of the two archetypes with the conscious personality. More precisely, a dialogue (symbolic transformations) must be maintained between personal consciousness and the contrasexual archetype. Although failure to integrate the two may result in psy- chopathology in both sexes (Stevens, 1983), Singer (1973) observed that women are disadvantaged in soci- eties that esteem and reward traits associated with the animus (individualism-autonomy, technical skills, clar- ifying facts), but which either forbid or prevent women from developing their animus and integrating it with their personal consciousness. Therefore, integration of the psyche is more likely to be impaired in women. Bem (1977) and Spence and Helmreich (1978) found that masculinity scores were positively associated with self-esteem in females, further suggesting that some degree of conscious integration of the animus is neces— sary for psychological well-being. However, the mean- ing of masculinity-femininity used by these investiga- tors appears more broadly defined than Jung's specific conceptualizations embodied in the anima-animus. The differences in the psychic structures of males and females are also reflected on two bipolar pairs of cognitive functions: sensation-intuition and thinking- feeling. Sensation and intuition are nonrational modes of perception which are rationally evaluated by think- ing and feeling functions. Sensation operates through the five senses and is tangible and concrete, whereas intuition is perception via unconscious means or pro- cesses. Thinking addresses the question of whether something is valid or invalid, logical or illogical; whereas feeling assesses whether something is important or unimportant, liked or disliked, valuable or worth- less (feeling means value judgements, not affect). The two attitudes and four functions combine to form eight distinct function types. Both Jung (1971) and Stevens (1983) observed sex differences in function type. Thinking and sensation tend to be more dominant in men whereas feeling (introverted feeling in particular) and intuition tend to be more dominant in women. Mosher and Reyher (in preparation), using the Singer-Loomis Inventory of Personality (SLIP) (Singer & Loomis, 1984), found that males scored significantly higher than females on extraverted sensation and extraverted thinking, whereas females scored signifi- cantly higher on introverted feeling. Females also scored higher than males on introverted intuition, but this difference was not significant. Similarly, using various measures, Hirschman (1983) found external sen- sation seeking (similar to extraverted sensation) and external cognition seeking (similar to extraverted thinking) significantly higher in males than females and positively correlated with masculinity, as measured by the Personal Attributes Questionnaire (PAQ) (Spence & Helmreich, 1978), in both sexes. Internal cognitive 10 stimulation, emotional involvement and imagery, on the other hand, were significantly higher in females than in males and positively correlated with femininity in both sexes. If attitude is in fact related to psychopathology, then function type, which includes attitude, also ought to be related to psychopathology. Function type is a more refined way of describing psychic structure that is embodied in such measures as the SLIP. If gender differences in function type do underlie gender differences in psychopathology, recognizing that social biases may exist in diagnostic criteria (Kaplan, 1983), then the following hypotheses can be tested: 1. Females will have higher scores than males for (a) introverted feeling and (b) introverted intuition. 2. Males will have higher scores than females for (a) extraverted thinking and (b) extraverted sensation. 3. Females, due to more highly developed intro- verted feeling and introverted intuition function types, will have higher measures of (a) somatization, (b) depression, (c) anxiety, (d) obsessive-compulsive symptoms, (e) phobic anxiety, and (f) interpersonal sensitivity (feelings of inadequacy and inferiority) than males. 4. Males, because of their greater degree of extraverted thinking and extraverted sensation function 11 types, will score higher than females on measures of (a) sociopathy, (b) impulsivity, and (c) hostility and (d) paranoid ideation, two symptoms often observed in personality disorders. 5. Females, because of their introverted feeling and intuition function types, will have higher measures of intensity of psychopathology than males. 6. Introverted feeling and introverted intuition will be positively associated with measures of (a) som- atization, (b) depression, (c) anxiety, (d) obsessive- compulsive symptoms, (e) phobic anxiety, (f) inter- personal sensitivity, and (g) intensity of psychopathology. 7. Extraverted thinking and extraverted sensation will be positively associated with measures of (a) sociopathy, (b) impulsivity, (c) hostility, and (b) paranoid ideation. In addition, if psychological adjustment in females is related to the degree of conscious integration of the animus (masculine component) with personal consciousness, then the following additional hypotheses should be supported: 8. For females, animus scores will be negatively associated with measures of (a) intensity of psychopathology, (b) somatization, (c) depression, (d) 12 anxiety, (e) obsessive-compulsive symptoms, (f) phobic anxiety, and (g) interpersonal sensitivity. Similarly, if the degree of conscious integration of the anima (feminine component) is related to psychological adjustment in men, then: 9. Anima scores for males will be negatively associated with measures of (a) intensity of psychopathology, (b) sociopathy, (c) impulsivity, (d) hostility, and (e) paranoid ideation. A secondary concern of the present study was to compare the Anima-Animus Scale, developed for this study on the basis of Jung's definitions, with the broader measures of masculinity-femininity used by Bem (1974) and Spence and Helmreich (1978). Method 313191102113 Participants in this study consisted of 120 college students from an introductory psychology course at Michigan State University. The group was made up of 55 males and 65 females. Each person received supplemental credit toward course grades for participating. Materials Singer-Loomis Inventory of Personality (SLIP). Singer and Loomis (1984) developed a new inventory, the Singer-Loomis Inventory of Personality (SLIP), to assess eight functional types or cognitive modes based on Jung's two attitudes of introversion and extraver- sion, and the functions thinking, feeling, sensation, and intuition. Conceptually, the SLIP offers a major improvement over other Jungian instruments such as the Gray-Wheelwright (Wheelwright, Wheelwright & Bueler, 1964) and the Meyers-Briggs Type Indicator (Myers, 1962) in that it uses scaled items rather than a forced-choice format. This lack of constraint allows opposite functions such as thinking and feeling to be 13 14 measured independently, whereas with the earlier instruments the choice of one of these bipolar oppo- sites automatically resulted in the rejection of the other. Items pertaining to each of the eight cognitive modes are assessed on Lickert-type scales (1 = never, 2 = occasionally, 3 = about half the time, 4 = usually, and 5 = always). Conversion of SLIP raw scores to percent scores creates an ipsative scale (Clemans, 1973). Ipsative measurements for each individual are distributed about the mean for that individual, not about the population mean. This precludes comparisons across a population of individuals and results in artifactual negative cor- relations between scale items. For these reasons, SLIP raw scores were used, despite the high desirability of the items and the possible creation of a negative pre- sentation of self by not endorsing an item. Notwith- standing these psychometric limitations, and the ques- tion of whether such self-report measures actually mea- sure Jung's multidimensional conceptualizations, the SLIP produced sex differences in function type in line with Jung's theory (Mosher, 1985). Additional evidence supporting the SLIP's validity is limited to one pre- dictive study (Loomis & Saltz, 1984), in which function type was used to successfully predict artistic style. Singer and Loomis (1984) reported internal reliability 15 coefficients for the eight function types ranging from .56 to .71, with an average alpha coefficient of .64. Symptom Checklist 20R (§QL'90-Bl- Psychopathology was measured using the Symptom Checklist 90R (SCL-90-R) (Derogatis, 1977; Derogatis, Lipman, & Covi, 1973). The SCL-90-R is a self-report inventory in which subjects are asked to rate how much they are distressed by 90 different symptoms on a 5-point scale ranging from 0 = not at all to 4 = extremely. The 90 symptoms are grouped into the following nine categories or types of psychopathology: somatization, obsessive-compulsive symptoms, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. In addition, the SCL-90-R provides a measure of the intensity or depth of psychopathology with the Global Severity Index (GSI). Derogatis (1977) has provided extensive evidence of the validity and reliability of the SCL-90-R, including test-retest reliabilities for the SCL-90-R subscales ranging from .78 to .90, and alpha coefficients ranging from .77 to .90. Anima-Animus Scale. This scale was developed to provide a measure of the degree of conscious integra- tion of the contrasexual archetypes; the anima and the animus. Jung (1959b) described the animus (masculine element) of the female psyche as an assertive, 16 autonomous, and discriminative function, whereas the anima (feminine element) of the male psyche was described as a capacity for emotion and relatedness. On the other hand, the unconscious or undeveloped anima and animus may be manifested in negative ways; the anima as rigidity, stubbornness, and moodiness, the animus in an aggressive, opinionated manner. Jung acknowledged his limitations in understanding the psy- chology of women (Mattoon, 1981). Consequently, he deferred to his female coworkers, especially Emma Jung (1957) and M. Esther Harding (1970), for presentations of some of the specific aspects of female psychology and the animus in particular. It seems clear that Jung's concept of the anima- animus is more narrowly defined and different from the more popular conceptualizations of masculinity-feminin- ity embodied in such instruments as the Ben Sex Role Identity (BSRI) (Bem, 1974) and the short form of the Personal Attributes Scale (PAQ) (Spence & Helmreich, 1978). The BSRI and the FAQ, incorporating such addi- tional items as loyalty, childlikeness, excitability, and need for security, appear to encompass broader for- mulations of masculinity-femininity than those proposed by Jung in his concept of the anima-animus. A11 20 items (10 anima, 10 animus) of the Anima- Animus Scale correspond closely to Jung's specific 17 definitions, and include both positive (integrated) and negative (unconscious or repressed) aspects of each archetype. A complete list of items, direction of scoring, and the corresponding references can be found in Appendix A. Each item of the Anima-Animus Scale is scored from 1 to 7 on the basis of how well it describes the individual. Total anima and animus scores (possible range 10 to 70) are obtained by summing the individual scores for the 10 respective items, with negative aspects being reverse scored. Minnesota Multiphasig Personality Inventory (MMPI). Sociopathy was measured using the Psychopathic Deviate (Pd) scale of the MMPI (Hathaway & McKinely, 1951). The 50-item Pd scale measures general social maladjustment, antisocial behavior, and the capacity for angry, impulsive acting out (Carson, 1969). A con- siderable body of research (Dahlstrom, Welsh, & Dahlstrom, 1982) has shown that the MMPI scales have construct validity for measuring personality traits in the normal range, and that the Psychopathic Deviate Scale (Scale 4) has a demonstrated test-retest relia- bility of .88. Impulsiveness Scale (IS). Impulsivity was mea- sured using the lO-item Impulsiveness Scale (IS) developed by Schalling, Edman, and Asberg (1983). The IS was developed from a series of item analyses of 18 «other impulsivity scales, including the Barrat Impul- siveness Scale (BIS) and the Guilford Zimmerman Temper— ment Survey (Guilford & Zimmerman, 1949). The scale uses a 4-point scoring system and has been used widely by Schalling, et a1. (1983) in their research on impul- sive cognitive style. The IS has demonstrated adequate reliability (a test-retest reliability of r=.7l) and is significantly correlated with other impulsiveness mea- sures, such as the Impulsiveness Venturesomeness Empa- thy (IVE) scale (Eysenck & Eysenck, 1978). gem Sex Role Inventory (BSRI). This scale con- sists of a list of adjectives (20 masculine and 20 fem- inine) which are endorsed using a 7-point scale.' Responses to the masculine and feminine adjectives are summed and averaged to yield a Masculine and Feminine scale score. These scales are based on empirical vali- dation with several undergraduate populations (Bem, 1974) and have been found to be highly reliable. Bem (1974) reported alpha coefficients for internal consis- tency of .86 and .82 respectively for the BSRI Mas- culinity and Femininity scales. Test-retest reliabili- ties of r=.90 for Masculinity and r=.90 for Femininity were also presented. Personal Atrributes Questionnaire (259). This is a self-report instrument consisting of 24 trait descriptions, each arranged on a bipolar scale (Spence 19 & Helmreich, 1978). The Masculinity scale contains eight items that are considered socially desirable in both sexes, but thought to occur to a greater degree in males. Similarly, the eight items for the Feminine scale are defined as desirable characteristics in both sexes, but exist to a greater degree in females. The Masculine-Feminine scale contains items for which one pole is desirable for males and the other pole is desirable in females. Low correlations between the PAQ scales and the Marlowe-Crowne Social Desirability Scale (Spence, Helmreich, & Stapp, 1975) support that the PAQ's validity is minimally distorted by a tendency for subjects to respond in a socially desirable manner. Alpha coefficients for the PAQ as reported by Spence and Helmreich (1978), were .85, .82, and .78 respec- tively for Masculinity, Femininity, and Masculinity- Femininity (M-F). esi The present study incorporated both correlational and experimental designs. The manipulation of the sub- ject variable gender, as in the present study, is clas- sified by some investigators as an instance of the experimental method (Wood, 1981), although the results of such experiments cannot be interpreted to make causal statements. Gender is a nonmetric, categorical 20 variable, and not an individual difference variable as described by Humphreys (1978). Therefore, the results of manipulating gender can be viewed with more confi- dence than the correlation of an individual difference variable, but they are not a guarantee of causality. Because of the exploratory nature of the present study and the difficulty of manipulating subject variables, such as function type, a primarily correlational design was selected as most appropriate. The rationale of this approach was to use correlation to discover possi- ble causal relationships which could then become the subject of experimental investigations in subsequent studies (Wood, 1981). Setting end Procedure Participants were recruited from introductory psy- chology courses by using sign-up sheets with the title of the experiment labeled: Male and Female Differ- ences. Some of the classes required student participa- tion in research projects, while in other classes par- ticipation was optional. The instruments were adminis- tered to small groups of participants, 15 to 20 per group, in a classroom setting. The average time required to complete the instruments was approximately 75 minutes. The primary experimenter (male) was aware of the hypotheses of the study, whereas the assistant 21 experimenter (female) was "blind" with respect to the purpose of the experiment. Both experimenters shared equally in all duties, with the exception that the pri- mary experimenter took the lead in describing the study and addressing the participants. The limitations of self-report instruments due to social desirability and impression management are well documented (DePaulo, Kenny, Hoover, Webb, & Oliver, 1987; Goffman, 1959). In an effort to enhance self- disclosure, group size was purposely kept small, and an effort was made to engage each group personally during the explanatory phase of the experiment, rather than fostering a solely didactic, one-directional relation- ship. The participants were asked to respond to each item with as much candor as possible. They were assured that there were no right or wrong answers, and anonymity was ensured by instructing the participants that no identifying information was to be written on the questionnaires. The standard consent information was given verbally, and the participants were instructed that the return of their completed question- naires constituted their informal consent to partici- pate in the experiment. At the same time, a concerted effort was made to control for unintentional expectancy effects, demand characteristics of the experimental setting, and 22 subject roles (Kazdin, 1980). This was attempted by using a "neutrally" worded instruction sheet that was attached to each packet of questionnaires. In addi- tion, the same carefully constructed outline was used to guide the introductory and instructional phases of the experiment. The primary experimenter made every effort possible not to convey demand information regarding the experimental hypotheses, while at the same time trying to elicit self-disclosure. The order in which self-report instruments are administered can have various effects on research out- come, including such factors as subject's self-atten- tion and public self-consciousness (Osberg, 1985). To control for these various effects in the present study, the instruments were administered in counterbalanced order by systematically rotating the order in which a given instrument occurred in the packets of question- naires. This procedure resulted in packets consisting of several different sequences of instruments, which were then randomly distributed to each group of partic- ipants. Results Means and standard deviations for the SLIP sub- scales are reported in Table 1. Although norms were not available for this research version of the SLIP, the means and standard deviations were well within the range of scores from a previous study by Mosher (1985). Table 2 presents means, standard deviations, and norms (if available) for all measures of psychopathol- ogy. Mean scores for all subscales of the SCL-90-R were considerably higher than the norms reported by Derogatis (1977). These norms, however, were from a significantly different age group (mean age = 46) than the present college-aged population. The means from the present study more closely matched the norms reported for adolescent outpatients. The mean score for the MMPI Scale 4 (Psychopathic Deviate) was also higher than the normative score reported by Dahlstrom, Welsh, and Dahlstrom (1982). Norms were not available for the Impulsiveness Scale. Means and standard deviations are presented by gender for the BSRI, PAQ, and the Anima-Animus Scale in Table 3. Mean scores for the BSRI and PAQ appeared to 23 24 .Ammmac umemozm mm.6 Hm.mv mm.e sm.ve cowuasocH emuum>muoxm hm.m mm.se em.e ms.ae :Oaummcmm emuom>auuxm mm.m me.eq mm.o mm.as maaawma emuum>muuxm sm.m mm.me mm.o m~.me assesses emuum>aupxm A~.A so.ae GN.A m~.ms ceabhsucH emuum>oubcH mo.e. mv.ms 4H.G em.mv coauamcmm emuuo>ouucH mm.m mo.>e om.m om.he mcflsmmm emuum>ouocH H>.w mn.mw Hm.m mm.mv mcflxsflcs pouuo>ouucH cm Immwm om Immmm mamom chosum msofl>oum macaw Hammond .mmuoom waoomnsm qum How Awpsum msofl>mum one Hammond. mQOHumw>mQ Unmocmwm use memo: H magma .Aamwuzv Ammmav Eouumanmo w .nmamz .Eouumazmo Scum at onom Hes: “Om mEuoz .maauz mucofiumauso ucwo Immfioea .64 I owe can: .snmnz smashedcoz .Aaamac maummoumo scum muomuqom ooh mauoza 25 I I OH.< hm.ma no.m Hv.mH wnquOAUOmaconomIHmzz I I I I mm.v oo.vH wamom mmocm>flmHsQEH mo. mm. am. an. em. em. .Hmov xoUCH >uauo>mm HMQOHU me. me. mm. as. on. om. smaoauonosma mm. mo.H «v. em. mm. mm. coflumopH owocmumm mm. hm. Hm. ma. on. me. aumwxc< cabana mm. o~.H ow. om. ma. mm. unassumom he. am. am. cm. on. mm. suwwxca an. em. as. em. em. -.H coammmudma mm. mo.H mm. am. mm. 6~.H muw> Iflufimcom accomuomwmucH me. am. me. am. we. NN.H m>amHsdeooIm>ammmmno as. Nb. we. cm. mw. mm. coHumNflumEom mIoaIAOm cm coo: cm coo: om coo: , onom mmucowummuso mmEHoz Aomanzv ucmommHOUc pcofiummcoz wosuw ucomwum .monom mmoHonuomonomwm wow .me02 one Mpsum ucomowmv mcofiumfl>oo oumocmum use memo: N manna 26 aAmhmHv SUflQHEHQE w QUCGQW EOHM mEHOG Ofim .Hoo. M.Ma«« .mo. v ma .mahnz .mmnuz .vhmav 2mm scum wagon Hmmmm I I I I mm. em.m me.p~ ms.~ qe.a~ sameHeca sausage I I I I em.HI mo.e HH.ms G6.e «.mv mseHc< I I I I H~.H om.m mH.om om.m mm.me ashes I I I I so. em.s om.m m6.a om.m mcsmouea< cam m~.q mm.~H NH.4 m6.6H IIImm.m GH.4 m~.HN ~4.m me.na mIz cam mm.m sm.a~ ma.m ms.- IIIHH.m Hm.m ~m.mm H6.m mm.a~ suHcH:HEmm can mm.e sm.ms ma.v me.Hm mm. I Gm.m mm.om mm.m mh.o~ suflcaasommz can I I I I om.H mm.s mm.m~ o6.m mm.s~ summOHec< Hmmm mm. Ho.m mm. 44.4 IIImm.e mm. 4H.m mm. mm.s muses:H5ma Hmmm mo. am.s he. sm.s Im~.~I he. as.v mo. Ho.m suflcflasommz Hmmm om m mm m IIMII mm m mm m mmmmm mmHmEam mafia: Amouzc 1mmnz. mmEHOZ mOHMEwm wwfimmi wosum ucmmoum a muflcflcflawmImuHsHHsommz Hem AmEHoz ocwiwosum usommumv msoHDMH>mQ pumpcwum cam memo: m OHQMB 27 fall within the range of reported norms (Bem, 1974: Spence, & Helmreich, 1978), with the exception of the PAQ M-F scores. Because the Anima-Animus Scale was constructed for the present study, norms were not available. Alpha coefficients of internal consistency for the Anima-Animus Scale were .54 for Anima and .50 for Animus. Although the alpha coefficients were some- what lower than those of the SLIP (average alpha = .64), they were not considered to be an important prob- lem at this point in the development of the Anima- Animus Scale. They were not critical insofar as the scale entering into significant relationships with the other measures. Furthermore, the slightly lower alpha coefficients were considered a fair trade-off to pre- serve the integrity of the conceptual basis of this Jungian derived scale. Table 3 also shows that females scored signifi- cantly higher on BSRI Femininity, PAQ Femininity, and PAQ M-F than did males. Males had significantly higher mean scores for BSRI Masculinity. The androgyny scores in Table 3 were calculated using a metric derived by Bobko and Schwartz (1984) to provide a continuous, rather than categorical, variable suitable for multiple regression analysis. The differ- ence method of deriving androgyny proposed by Bem (1974) has been criticized (Spence, Helmreich, & 28 Strapp, 1975) because a person scoring M = 7 and F = 6 would receive the same androgyny score as someone who scores M = 2 and F = 1. While the median split method advocated by Spence and Helmreich (1978) avoids some of the problems of the difference method, it has not com- pletely filled the need for an adequate measure of androgyny. The dichotomization of the original mas— culinity and femininity scores may cause a loss of a great deal of information (Lubinski, Tellegen, & Butcher, 1981, 1983), and the resulting categorical androgyny variable is not easily amenable to multivari— ate correlational analysis. Therefore, Bobko and Schwartz (1984) developed a metric ([(K-l) - |M-F|] x [(M+F/2], where K = maximum scale value) that they believed better incorporated the theoretical assump- tions of androgyny and provided greater statistical utility. In the first factor, the absolute value of the difference between M and F was preferred over the algebraic difference since the index thereby focused on the dimension "androgynous" - "not androgynous". The first factor was weighted by the average of the M and F scale scores, thus separating androgynous from undif- ferentiated individuals by guaranteeing that more weight was given to persons scoring high on both sub- scales than to persons scoring low on both. Using the 29 BSRI, the authors obtained a test-retest reliability of ;=.88. W The manipulation of gender was tested by univari- ate analyses of simple effects using one-tailed t-tests. The one-tailed test was selected because of the strong theoretical and empirical bases of the hypothesized relationships. Wis—l. Hypothesis 1 predicted that females would have higher scores on the SLIP than males for (a) intro- verted feeling and (b) introverted intuition. Both hypotheses 1a and 1b were corroborated (see Table 4). Female (M=49.82) and male (M=45.11) means differed sig- nificantly (r(ll8)=4.18, 95,.001) for introverted feel- ing. Similarly, female (M;49.52) and male (M=46.75) means also differed significantly (g(118)=2.12, pg.05) for introverted intuition. Hypothesis 2 It was expected that males would have higher scores on the SLIP than females for (a) extraverted thinking and (b) extraverted sensation. Hypotheses 2a and 2b were not corroborated (see Table 4). Male (M=43.13) and female (M=43.43) means did not differ 3C) Tab1e 4 Cgmpiigson of Mean Function Type and Mean Psychopathology78cores Between Males and Females _§ema1es (N-GS) Males (N-SS) Instrument/Scale X _§2_ i _§2_ t ergg Introverted Thinking 48.98 7.20 47.47 6.28 1.21 Introverted Feeling 49.82 5.70 45.11 6.63 4.18"* Introverted Sensation 50.08 6.14 47.39 5.33 2.44+ Introverted Intuition 49.52 6.81 46.75 7.55 2.12' Extraverted Thinking 43.43 7.06 43.13 5.98 .25 Extraverted Feeling 46.09 6.78 42.36 6.42 3.07++ Extraverted Sensation 41.52 6.38 40.67 6.74 .72 Extraverted Intuition 45.29 6.66 43.21 7.17 1.64 SCL-90-R Somatization .89 .60 .65 .14 2.11' Obsessive Compulsive 1.23 .80 1.21 .76 .17 Interpersonal Sensitivity 1.30 .84 1.21 .81 .59 Depression 1.32 .86 1.11 .81 1.37 Anxiety .91 .71 .78 .68 1.01 Hostility .80 .68 .88 .79 -.60 Phobic Anxiety .53 .75 .40 .63 1.02 Paranoid Ideation .97 .75 .92 .76 .36 Psychoticism .83 .69 .76 .71 .49 Global Severity Index (GSI) 1.01 .64 .86 .64 1.22 Impulsiveness Scale 13.54 4.86 14.64 4.23 -1.31 MMPI-Scale 4 Sociopathy 17.61 5.00 19.35 5.03 -1.88' ’2 1 ...2 : +2 5 .05, ** g .01, .05, one-tailed. .001, one-tailed. two-tailed. two-tailed. 31 significantly (r(ll8)=.25, pg.40) for extraverted thinking, nor did male (M=40.67) and female (M=41.52) means differ significantly (§(118)=.72, pg.24) for extraverted sensation. Hypothesis 3 Females, due to their greater predicted degree of introverted feeling and intuition function types, were expected to have higher scores for (a) somatization, (b) depression, (c) anxiety, (d) obsessive-compulsive symptoms, (e) phobic anxiety, and (f) interpersonal sensitivity on the SCL-90 than males. Although mean differences were in the predicted direction (see Table 4), female (M=.89) and male (M.65) means differed sig- nificantly (r(118)=2.11, pg.05) only for (a) somatiza- tion. Recognizing the equivalence of the t-test and correlational analysis allowed the use of partial cor- relation to further investigate the relationships of these measures of psychopathology to gender, while con- trolling for introverted feeling and introverted intu- ition, which were related to gender. However, none of the resulting partial correlations were significant (pg-05) - mine—SEA Males (M=19.35), as predicted, scored signifi- cantly (p(118)=-1.88, pg.05) higher than females 32 (M=17.61) on (a) sociopathy as measured by the MMPI Scale 4. Males (M=14.64) also had a higher mean score than females (M=13.54) for (b) impulsivity, as measured by the Impulsiveness Scale, however this difference was not significant (r(118)=-1.3l, pg.09). In correlation terms, r(.12) was also not significant (p5.09), however correlational analysis allowed the use of partial cor- relation to further investigate the relationship between gender and impulsivity, while controlling for introverted feeling and introverted intuition. As a result, impulsivity was found to be positively associ- ated (r=.16, pg.05) with being male, thus corroborating hypothesis 4b. It was expected that males would have higher scores on the SCL-90-R than females for (c) hostility and (d) paranoid ideation. Hypotheses 4c and 4d were not corroborated (see Table 4). Male (M=.88) and female (M:.80) means did not differ significantly (p(118)=-.60, p5.27) for hostility, although the dif- ference was in the predicted direction. Male (M=.92) and female (M=.97) means did not differ significantly (r(118)=.36, p5.35) for paranoid ideation. Partial correlations, controlling for introverted feeling and intuition, were not significant (p5.05). 33 M5 The prediction that females, because of their introverted function types, would have higher measures of intensity or depth of psychopathology on the SOL-90- R than males, was not corroborated. Table 4 shows that the mean Global Severity Index (GSI) score for women was 1.01 versus .86 for men. Thus, although this dif- ference was not significant (p(118)=1.22, p5.11), the mean difference was in the predicted direction. The partial correlation between gender and GSI, while con- trolling for introverted feeling and intuition, was not significant (pg.05). Correlational Findings The foregoing analyses indicate that, as predi- cated, females displayed greater degrees of introverted feeling and intuition function types, and that females also tended to have somewhat higher scores than males for symptoms of neuroses, depression, and intensity of psychopathology. On the other hand, males did not score higher than females on extraverted thinking and sensation as expected. However, males did score significantly higher than females on measures of sociopathy and impulsiveness as predicted. The next step in the progression of analyses was to examine the associations among psychopathology and 34 two aspects of Jungian theory: the function types and contrasexual archetypes. Therefore, correlational analyses were used to test these predicted relation- ships (hypotheses 6 through 9). nypotpesis 6 As expected, Table 5 reveals that correlations (r/r) for introverted feeling/introverted intuition, as measured by the SLIP, were significantly (p5.05) and positively associated with (b) depression (.25/.29), (c) anxiety (.24/.3l), (d) obsessive-compulsive symp- toms (.19/.24), (e) phobic anxiety (.18/.28), (f) interpersonal sensitivity (.24/.29), and (g) intensity of psychopathology (.24/.30), as measured by the SCL- 90. However, hypothesis 6a (somatization) was not cor- roborated, although the correlations (.12/.12, pg.09) were in the predicted direction. The foregoing findings are consistent with the hypothesized linkage between introverted feeling, introverted intuition, and psychopathology. Moreover, the non-significant findings for introverted thinking, introverted sensation, and psychopathology are incon- sistent with alternative explanations based on general- ized response biases. Impression-management (DePaulo, Kenny, Hoover, Webb, & Oliver, 1987: Goffman, 1959) provides an alternative explanation for this pattern of results. It is possible that the items constituting I35 .Hoc. M m ace .no. N N as .mc. v N a Ho.I ~o.I mo.I ~H.I cma. 4h~.I no. we.I azueno«oom v ouuomIHazs nu. ac. cc. mo.I mo. Ho.I mo. cmH.I uueom oaoco>uuusdsn and. He. mo. mo.I «econ. Ho.I 4¢v~. No. .Hmo. xoUcH auuuo>om decode «ma. do.I oH. nc.I ¢.v~. vo.I cod. Ho.I Endowuocoaaa «a. Ho.I ~c. mo.I ccnw. n~.I «ca. ~o. sawueocu cuoceuem ma. co. co. mo.I 4.4om. so. .63. No. suoaxc< canons as. 66.I mo. sc.I .ma. .H.I so. .o.I suaaauuoz «ma. mo. AH. me. accan. so. ccvw. ac. sawdust ca. ~o.I mo. wo.I «ccom. co. ccmw. ma. codenoudoc «as. Ho. mo. MA... eecmN. mo.l cavw. No... Tar/«H Iamcom occuuoauoucu 4mm. Ho. as. Ho.I ..q~. so.I 4mg. Ho.I m>auflaaeooIosauuumno we. so.I vo. mo.I ~H. Ho.I NH. mo.I casusuwueeom mIoaIaum :OMuwsucn :OMummcmm mafiaooh mcflxcfizs cofluwzucn sewucmcmm mcfiamom Ucfixcwne mnmomquOESuumCa cauuo>uuuxm pouuo>euuxm pouuo>euuxm oouuo>msuxm pouuo>0uucm couuo>ouucn oouuo>ouucm couuo>ouucn 4Nm0~ocuemocoxwa one onus newuocsm mo mamaamc< Hocofiueaouuoo ucoeoz uODOOum coaucom a «sane 36 the introverted feeling and intuition scales are inher- ently less desirable than introverted thinking and sen- sation scale items. To test this possibility, ten randomly selected introverted feeling or introverted intuition items, and ten introverted thinking or introverted sensation items were randomly paired. Twenty participants from the same pool of college students were asked to select an item from each pair of items that they thought a healthy, well adjusted person would most likely select. If there was no difference in social desirability, then it was expected that participants would not favor either member of a pair. Table 6 presents a tabulation of item choices and the resulting chi square test for each matched pair of items. As Table 6 shows, none of the chi square tests for the ten matched pairs of items were significant (p_<_.05) . Hypothesis 7 Contrary to expectations, Table 5 shows that cor- relations (r/r) for extraverted thinking/extraverted sensation, as measured by the SLIP, were not signifi- cantly (p$.05) associated with (a) sociopathy (-.12/- .02) as measured by the MMPI Scale 4. Similarly, extraverted thinking/extraverted sensa- tion were not significantly (p5.05) associated with (b) 37 N. HA m OH m. NH m m w.H ma n m N. m HH 5 m. w NH w N. ad a m N. a HH v m. NH m m 0.0 OH OH N N. HH m H IIIMIMW cofluflcucH Umpum>0uucH cofluwmcmm cmuuw>ouucH “03552 Mama Nx \mcflaoom Umuum>ouucH \mcflxcflca Umumm>ouucH cocoumz .mEmuH mo mused cocoumz coauflsccH cmuuw>0uucH \mcflaooh couno>ouucH 0cm cONummcom cmuuo>ouucH\mcchflcB cwuuo>onucH mo comwnmmfioo m OHQMB 38 impulsivity (-.08/.09), as measured by the Impulsive- ness Scale, nor were they significantly (pg.05) associ- ated with (c) hostility (-01/-.06) and (d) paranoid ideation (-.09/-.01) as measured by the SCL-90-R. Therefore, hypotheses 7a through 7d were not corrobo- rated. Mes—1&9 As expected, the Animus scores, which are not to be confused with introverted function types (Hypothesis 6), for females (see Table 7) were significantly and negatively associated with (a) intensity of psy- chopathology (r=-.29, pg.01), (b) somatization (r=-.23, pg.01), (c) depression (r=-.23, pg.05), (e) obsessive— compulsive symptoms (r=-.24, p5.05), and (g) interper- sonal sensitivity (r=-.31, pg .01) as measured by the SCL-90-R. Female animus scores were also negatively associated with (d) anxiety (r=-.19) and (f) phobic anxiety (r=-.20), although these correlations were not significant (pg.05). Hypothesis 9 As predicated, the Anima scores for males (see Table 7) were significantly and negatively associated with (a) intensity of psychopathology (rs-.29, pg.05), (d) hostility (r=-.37, p5.01), and paranoid ideation (r=-.27, pg.05) as measured by the SCL—90-R. Table 7 Pearson Product Moment Correlational Analysis of Anima-Animus Scale and Psychopathology. 39 Females (N-65) Males (N-SS) Instrument/Scale beige Animus Agimg Animus SCL-90-R Somatization -.12 -.23' -.26' -.33" Obsessive-Compulsive -.18 -.24* -.26' -.48"' Interpersonal Sensitivity -.31" -.31*' -.31" -.38" Depression -.29" -.23' -.25‘ -.37" Anxiety -.08 -.19 -.24' -.38" Hostility -.23' -.30" -.37" -.23‘ Phobic Anxiety -.06 -.20 —.30' -.39" Paranoid Ideation -.24' -.24* -.27‘ -.21 Psychoticism -.27' -.25' -.26' -.28* Global Severity Index (GSI) -.23* -.29‘* -.29' -.36“ Implusiveness Scale -.08 -.22' -.16 -.49"* MMPI-Scale 4 Sociopathy -.27* -.39"‘ -.341" -.21 'p 1 .05. *‘p i .01. *"p i .001. 40 Anima scores for males were also negatively and significantly associated with (b) sociopathy (r=-.34, p5.01) as measured by the MMPI Scale 4. However, Anima scores for males were not signifi- cantly (pg.05) associated with (c) impulsivity (r:- .16), as measured by the Impulsiveness Scale, although the correlation was in the predicted direction. WWW At this point in the analysis, the significant findings relating introverted feeling and intuition, the contrasexual archetypes, and psychopathology were brought together, along with more traditional measures of masculinity, femininity, and androgyny, in an effort to evaluate their contribution to a predictive system for psychopathology. The notion that androgyny is an index of adjust- ment and psychological health (Bem, 1974) has received very little empirical support (Long, 1986). Lubinski, Tellegen, and Butcher (1981, 1983) have represented androgyny as the unique contribution of the linear by linear interaction, MXF. They argued that androgyny has utility only if the MXF term significantly added to a prediction system, over and above the main effects of M and F. Using hierarchical regression, the authors found that androgyny did not significantly add to the 41 prediction system beyond the main effect of M for posi- tive and negative affectivity. In those instances where androgyny did contribute, its association was in the direction opposite of that predicted. Using the SCL-90-R Global Severity Index, as rec- ommended by Derogatis (1977) when a single summary mea- sure of psychopathology is required, the present study attempted to test whether any of the three androgyny measures significantly added to the prediction system, in the same manner as Lubinski, et a1. (1981, 1983). In addition, the androgyny metric proposed by Bobko and Schwartz (1984) was compared with the linear by linear interaction, MXF, to test the author's contention that it contained substantially different information. In the present analysis, introverted feeling and intro- verted intuition were also included to test their con- tribution to the predictive system. Table 8 presents the hierarchical regression anal- ysis for the BSRI, PAQ, and Anima-Animus Scale on the dependent variable Severity Index (GSI). The main effects for masculinity were significant (pg.01) in all three cases. The main effects for femininity (anima) were only significant for the Anima-Animus Scale. In all three analyses, the main effects for introverted feeling were significant (pg.05), as were the main effects for introverted intuition for the BSRI and 442 ..onH. Iflsucm oeuuo>ouacw I an .ocuaoou pouuo>0uucu I an Nausea nu :Esdoo >uo>o cw auuco umuuu ecu .aszh Hmo pee o~nowus> use» new cauueaouuoo defiance on» aucomoudou u «caused pouonsa cannon use .ONalz .Ae uo .oxnon an cauuos 0:» mafia: nououm >c>00ucco any one oeuoc< mean can 00upc< ouacacaeou I m .>»«cw~:0ame I : .cOMucHouuoo ouaeaa oz» .cofiussqo 0:» :« modnsaus> ocficououa ecu was new oc«-0uucoo nouuo .OuspoOOua 056a ecu mama: saucy coquoeuoucu on» vcwuouco not» can couzmou deucososocu so am uuoouue cans ocmuouco an vocueuno one: cusp omega “mecz .Hoo. W Nose .uo. H We. .ma. w m. co. 5v. No. 2H 8 xom co. nv. No.I 2H x xwm co. mv. mo.l 2H x xom AH ~o. be. NO. ha x x0m cc. nv. ac.I hm x xwm co. mv. ho.I bu x xum cu co. me. cc. nawc< x xwm co. me. co. m x xom No. mv. co. m x flaw o co. vv. Ho.I QSENS< u now ac. nv. HA. 2 x xmm es. ov. co. 8 x xvm o oo. .4. mc.I .uouec4 .ucau so. as. No. .ooaeca 04a co. .4. ms. scsmouec< Mama 5 co. vv. no.I oEwc‘uan=E«:< do. He. mo. bx! cc. vv. mc.I bx: w co. we. nc.I 30m co. dc. oc.I xwm oo. vv. mo. xom m «me. we. 0N. 2H Ho. ow. ma. 2H «no. vv. mu. 2H v sec. on. NN. hm «no. mm. oN. an «no. ov. ad. km N coco. cm. mN.I unauct oo. mm. ac.I h Ho. cm. as. h N came. vN. vN.I uEwc< ocean. mm. mn.I 2 caaNH. vn. om.I ! A mumMHMIMM Iml u defiance cansgue> wmmwzo Na IMI u deduced ofinmaus> ommmco N Iml u Hosanna ounsuuo> moum oamom assachmewcc omm HmnoHo mIomIaom co unawuumuoucH measuwme ccm «sway cofluocsm14uoccoo .NcNUOuccd .NJNCNCAEQL .Nuwc«~:0mmz uo ncOMnmoumoz ounce 43 Anima-Animus Scale. None of the interactions were sig- nificant (pg.05). The MXF interaction did not significantly add to the prediction system, over and above the main effects for masculinity. In none of the analyses did the androgyny metric of Bobko et a1. (1984) add to the pre- dictive system beyond the contribution of the MXF interaction, which was negligible. Other Findings Although not anticipated, female (M=50.08) and male (M=47.33) means differed significantly (p(118)=2.44, p5.05) for introverted sensation. In addition, females (M=46.09) also scored significantly (p(ll8)=3.07, p5.01) higher on extraverted feeling than did males (M=42.36). In regard to function type and psychopathology, introverted intuition was found to be positively asso- ciated with hostility (r=.19, pg.05), paranoid ideation (r=.23, pg.01), and psychoticism (r=.24, p5.01) as mea- sured by the SCL-90-R, and with sociopathy (r:.18, pg.05) as measured by the MMPI Scale 4. Also not anticipated was the positive association between intro- verted feeling and paranoid ideation (r=.16, p5.05) and psychoticism (r=.l6, p5.05). 44 Female Animus scores were found to be negatively associated with SCL-90-R measures of hostility (r=-.30, p5.01), and psychoticism (rs-.25, pg.05), Impulsiveness Scale measures of impulsivity ( =-.22, p5.05), and with the MMPI Scale 4 measure of sociopathy (r=-.39, pg.001). Although not hypothesized, female Anima scores were negatively associated with overall psy- chopathology (r=-.23, p5.05), interpersonal sensitivity ( =-.31, pg.01), depression ( =-.29, p5.01), hostility (r=-.23, pg.05), paranoid ideation (r=-.29, p5.05), psychoticism (rs-.27, p$.05), and sociopathy (r=-.27, p_<_.05) . For males, Anima scores were found to be nega- tively associated with somatization (r=-.26, p5.05), obsessive-compulsive symptoms (r=-.26, pg.05), inter- personal sensitivity (r=-.31, p$.01), depression (r=- .25, pg.05), anxiety (rs-.24, p5.05), phobic anxiety (r=-.30, p5.05), and psychoticism (rs-.26, pg.05). Similarly, male Animus scores were negatively and sig- nificantly (p5.05) associated with all measures of psy- chopathology, except paranoid ideation and sociopathy (see Table 7). gomparison of Masculipity-Eemipiniry Meespres One of the secondary purposes of the present study was to compare the Jungian based Anima-Animus Scale 45 with the BSRI and the PAQ. The intercorrelations for all three measures are presented in Table 9. The Anima subscale was positively associated with both BSRI Femininity (r=.41, p5.001) and PAQ Femininity (r=.41, pg.,001), and the Animus subscale was posi- tively associated with both BSRI masculinity (r=.34, p5.001) and PAQ Masculinity (r=.28, pg.001). Thus, there was support for the construct validity of the Anima-Animus Scale. There were, however, some differences between the Anima-Animus Scale and the other two measures. The Anima and Animus subscales were positively associated (r=.49, pg.001), indicating some overlap in the mea- surement of the two constructs, whereas the Masculinity and Femininity subscales were independent by design within both the BSRI (r=-.08) and PAQ (r=-.00). Simi- larly, BSRI Masculinity and PAQ Femininity were inde- pendent (r=.08), as were BSRI Femininity and PAQ Mas- culinity (r=-.03). The Anima subscale, on the other hand, was positively associated with BSRI Masculinity (r=.l7, p$.05) and PAQ Masculinity (r=.3l, p5.001), whereas the Animus subscale was not significantly (p5.05) related to BSRI (r=.12) and PAQ (r#.11) mea- sures of femininity, thus offering additional support for the construct validity of the Animus subscale. 446 .Hoc. M Mace .do. W Nae .mo. v No I «a. cacam. ac. «cm. ocean. Nc. cc. cc. cod. hI: OHmHsmEH mo. ast.I HO.I eccnn.I MH.I NH. «cavm.l eaaNm.I seenn.I «CVN.I .HmU. XHUCH auHuo>om HmnOHo mo. «NH.I ho.I ach.I mc.I mo. aHN.I eceNn.I ecehN.I ecmN.I EOHO Iauozusaa no. 40H.I ac.I camN.I ao.I cc. «INN.I accoN.I a¢NN.I camN.I :OHuuwcH UNOCQH NE N0. aan.I No.I caemN.I ceoN.I NH. asamn.I saNN.I secaN.I amH.I aumecfi oHnocm as. .sH.I .~.I .H~.I 4~.I as. .mH.I ...~n.I ..m~.I ...cn.I masseuse: no. aamN.I co. cacmn.I cmH.I coH. «acbm.l caNN.I eeamN.I vH.I xuucht no. caavn.I vo. eschv.l MH.I NH. «camv.l eeeom.l eeeon.l aemN.I :OHunOuQOD no. aaaNm.I no.I accvv.I aaH.I mo. acacv.I caacv.I scamn.I asaOM.I NUH>HuHmc0m HscoouwauoucH mo. 60H.I No. aeson.l OH.I NH. «econ.l sasvn.l sccvn.I aeHN.I 0>HmH5QEOU IO>Hmu0mno vc. NH.I mo. 66H. HH.I OH. caaom.I «QH.I eeecn.l eoH.I :OwuflnwumEOm mIchHUm LI: chmoupc< NuHchHEom NchHHsomm: anam0upc< NuHCHcHEom NuHcHHsoms: acamoupc< usEHc< cEHc< meum 04a o an axuoz pouooauou n.0cso .0 .0 o» eunuch 30 .ouoz .nn .uoa .--azo. .nn .uoa .Hanmzo. .ouoou onuo>ou. .nn .uoa .HH-a:o. .mom .uoa .ouso. .nm .uoa .unuazo. .moo .uoa .ozo. .onaa .quuuoz. .ouooo ouuo>ou. .nm .uom .Hu-azo. .bmaa .ucan .u. .ouOOo ouuo>ou. . .unm .Hnozu..uuouo mauo>ouv .mN .uua .Huumzo. .smmu .mcan .m. .n .uom .Hnmnu. .ouooo onuo>ou. .nn .uaa .HH-mzo. .vn .unm .Hucmzu .ouooa onuo>ou. .moo .uoa .wzo. .emma .ucsu .m. .aw .uom .Huumsu. .ouoon ouuo>wu. .smad .ocan .u. .nn .uoa .~H-a3o. h o m onus mNdzut onus mane uuoeu< no manan< aauaam: couuo ocoqnwooc onus on: ao>uuocuouua ouo=~o>o cu o~n< uuozuo Ou acououuuc:H o>uuoo~uom nouaunou mcuuuuucmuowuwo can ocunuwsmcuuouo uo canoomo o>auuucom amouuoOu uo chord uuonuo photo» o>uoaouuo¢ aocnauca no «apnoea uuozuo announce nooHoon ouuzonuuom uuozon :30 «.0:0 co cOwuou oxou 0» want scoot ouozuo Ou unaccouoaou uo macaw cocoa:o«uaudom o>auoonno can Hoodoo; - o>uuoouo couo:o«cumo auudwna dugouooooq OHQSuO 0U COMUGHGH Cd §H03 .cN .an .ou .5H .0H .ma .vd .nd .Nd .«H .cd .a .0 .h .w .m .v .n w .H .u:§«:¢. .on«:¢. .aaaa:¢. .uau1=¢. .oa1=<. .oeace. ..=¢«c¢. .oa1=<. .uaad=<. .an«=¢. ..=u«c<. .uaauc¢. 1.31:4. .oawce. .oach. AuQHMC‘. Aoawct. .osasca. .naauc¢. .oasc¢. m v n w mans wane mane noefiumeom >~u~¢0wooooo Condom wane uo>oz “MOEH‘ HO HQ>UZ .uuonusoa nonuuouoc oucoeououu ocwonHOu on» «0 zoom Ham: so: caucuosu 0.00am .oauou ucwoaah 0:» mafia: odoum noiacsnoEwc¢ < Xuccocat HICHEGQN STRTE UNIV. LIBRnRIEs ‘IIIHWIWWWINIIWIWIHIIWIWI“WWI 31293005330414