. .mmnMvr fizyc 4.. .vfifimfi . .V . . . _ V V ., . r l‘ 5.. am... . .a? A. 2?. V p . «1,: . («My 1.me hm“ . .w V . , 2 t. c. . .... . V :l?. V...a§i~ .V j u, . , , . . .V . . . .M‘wflx‘» . . . . V V V. fl.....n..V....u... . . V V .. . . amuka. V . Law... . V . x . 521.}! _ y ., , , . Va», 5... V m» V . . .V . . . .61“...Iv. \ . ., r . v v . . v . .. . .. u . ., . . _Wfifi.wnfiuw. A .ILarvt x. a . . ‘ VJ”. €31. . . . . , ‘. .. .. ,. .. . . , . A . .. V . , . .3 Ea V... .. V .. V. .. .. .V » fiCwW . . 9.5... . . V . . . _ . ‘ . $4.??? V .. . . Shufiurym . :tfil‘ - .01! .M .2. . . . ,H . . .V , ‘ V . 4 . Q 1 . a u...“ .mrrt. . V. V. 1.... . .2 im . yam“,- ui . . ., ._ V V Er . x. I A $2.... mm .7 . V ‘ afifi...£§nzma LVV V .r . V V . . . . ‘ $1..de .L . . . . . . .m: x .. 4 V U V. , V . 9...: . A . ., . , , . , 43.3.2 . .I 3...... . V : 0.x. . 1... 13.1. l cutout.- - _ 7 vflawvvvwfivu , . .. . . 3 .5. . , V ‘ V . . 1V .V y . . V . . . . @936 . . V. V . . .0. v)? V ‘ ‘ . . V . . . V , . V . . . .hll . . , ‘ , . z . , . V .2 . . m . v . . V . QRIQI‘W‘ . > . . V V V . . . . _ .13.)! u. #1 , . . . . ‘flvhflu . _ . ,V «Fl..- this.” Ito»..!att.tbx.ntit- 4w . . a ,. :. ..\ a V V. .. .3233. u ‘4. . I..'.‘£¢, V. r . V‘ "a . . ‘ u ' I . . . .. . al- . .v . . ‘ . 5“: I .9”. ~ . A o in! '0‘.“ ~ ’. ’l‘ in I'll . .31! 2 . L..- . . . \ V.....V..... . .. . Y.‘ Y. . . V . V .. 4|!”qu . Avl ‘ .“~On . t "l4l ' “D U ‘fihilluuufivfin {:4 .120! 19.; ..L. JWH‘JJS‘. .th’fl filly»... I\v( 1‘. I n “‘I1ll’>\i'|.l - ‘Rlvlltiu..7 4150‘»... .I 8, .\ t? 1.01 , ‘ x 3 r,‘ v... Mn . r. flux»! . \«V... V . Q0. ,IA:H.H1£QI . . . V)““» v. A. 7'. 3. b .. nc x. v? t... c. _ . . ‘ :1. . ...V i... _ ,1”. tours! . V 111‘ i . . x6151? V . r? V . 4.10 V . 1031' M . . {iii}; . . .1: V : . V . .134 nha_.a\...x.ix ‘x;.. .3-.- . .0... .\ ‘no. navy... ‘.I"I -‘i lo'tw I" .-... . .O‘OMSI VI): ull)‘ >Vt....fz....n. {1.1 -3: . V . ‘ {$1.}: .~ ‘ ‘Ivn...75.v\\‘l {WNW-9.x.” 1:: 1.4,, \«U1 . . 544 n. I ‘ 4 | 0‘ 5‘4 Alnlww' .3. 1.. In, 2... IklellvOli t ..v LN. I H1.-. 1th.}. 3‘11 .9 V 2 W 2.. {Is «I \r IV L}. V V s . zkltif: . ..V V 3. . V V . . . ‘ . .mxvrlus \Insln. .0 ‘ .v. 3.312% a. :1 . t .n. : . « $.16}: A .a.»....> 3 .9?! :3 24m :9. 19.x). is 1.1.291. 3V Ln. :ba»... .§.v.v.....:r-1:1\ . . . v1? ., 9., 0 It... .th‘U..-WVLQJ..I.~JWH'\I . vplizz. I’YVvvbfll .; , .. V «Hath... 1.....‘2 . . V . . . . . ‘ ‘ .. . 62:11. 1 l. »|\4.‘vt , l l. .. v" r. .bvvilxvllv v.05!) l u ulk ..u.4..¥Q1 V...V1». ticnlélsxl, f5 ‘ . V r .1. Lin!“ \ V . u: wruwdn,4...o.JW...;n. nun .Vlr ( 53; Imillbilflwlmulwmlwul so». '2 ‘) This is to certify that the dissertation entitled Unconscious Gender Priming and Perceptions of Women as Child Molesters in a sexually Ambiguous Situation presented by Grace Christine Gibson has been accepted towards fulfillment of the requirements for Ph-D- degree in _ __Es¥cho_lng y W v mjor professor Date 631‘ '7 in MS U is an Affirmative Action/Equal Opportunity lun'nm’on 0-12771 LIBRARY Michigan State Unlverslty PLACE IN RETURN BOX to remove We checkout from your record. TO AVOID FINES return on or bdore date due. DATE DUE DATE DUE DATE DUE _ll i-_____ Jail" Di L_J= usu leAnN’flnnetlve ActioNEquel Opponunltylnetltulon m ulna-9.1 UNCONSCIOUS GENDER PRIMING AND PERCEPTIONS OF WOMEN AS CHILD MOLESTERS IN A SEXUALLY AMBIGUOUS SITUATION BY Grace Christine Gibson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1996 ABSTRACT UNCONSCIOUS GENDER PRIMING AND PERCEPTIONS OF WOMEN AS CHILD MOLESTERS IN A SEXUALLY AMBIGUOUS SITUATION BY Grace Christine Gibson Researchers have speculated that many more women sexually abuse children than is currently documented, but the behavior is misperceived as "normal child care," and the sexual nature of the activity is "masked." Studies have demonstrated that men and women are judged differently when sexually abusing a child; women perpetrators are perceived‘ as less harmful and as less representative of child sexual abusers. This study tested whether implicit gender stereotypes underlie these differential perceptions. Implicit or unconscious stereotypes effect how a person perceives and judges situations without the individual's awareness; they are easily activated and difficult to counteract. Participants' (N=464; 119 males and 345 females) unconscious stereotypes of male aggressive behavior, female nurturant behavior, and male and female heterosexuality were iactivated using a priming method. Findings indicated that unconscious stereotypes of male aggression might best explain the differences observed in ratings of male and female targets. Priming with stereotypes of male heterosexuality had no influence on subsequent ratings of targets. In general, women targets were rated more positively than male targets; qualitative results suggested that the possible sexual aspects of the female targets' behavior were ignored. No support was found suggesting that womens' sexually ambiguous behaviors were judged as more nurturing than males' sexually ambiguous behaviors, although this hypothesis was inadequately tested. A replication of this study with methodological changes is necessary to substantiate the above findings. This dissertation is dedicated to the four women who have contributed the most to my development as a women, psychotherapist, and researcher: my grandmother - Katharina Lina Bender my mother - Brigitte Marianne Bender and two of my mentors - Teresa Bernardez, M.D. and G. Anne Bogat, Ph.D. Thank you all, for this work would not have been possible without what I received from each one of you. This dissertation is also dedicated to my husband, Paul Eric Luikart, and to my father, Gordon Seal Gibson. iv ACKNOWLEDGEMENTS Thanks to Bobbie Rosencrants and Angie Hoogterp for introducing me to the research on female sexual offenders; to my chair, Anne Bogat for her patience in guiding me through this process; to my undergraduate research assistants, Lynne Cessante, Rachel Duditch, and Devon Price for helping with the arduous task of data collection, and to Lynne and Devon for helping with everything else; to the ' members of my committee, Bob Caldwell, Galen Bodenhausen and Jackie Lerner; and to Beth Sipple Janick for being a true friend. TABLE OF CONTENTS LIST OF TABLES.................... .............. .....viii LIST OF FIGURES..... ......... .. .................... ....ix INTRODUCTION................... ................... ......1 Childhood Sexual Abuse ...................... .......3 Female Sexuality.............. ....... .............18 Implicit (Unconscious) Gender Stereotyping........24 Rationale of the Study...... ................ ......38 Hypotheses of Present Investigation ........ .......41 METHOD.. ............ . ........... . ................... ...47 Overview. ......... . .............................. .47 Subjects.... .................................. ....47 Material.... ........ ... ....................... ....48 Procedure ...................................... ...54 RESULTS........................ ................ ........58 Overview....................... ....... ............58 Results for the Aggressive Prime Condition........60 Results for the Nurturant Prime Condition.........61 Results for the Male Heterosexual Prime Condition... ......................... ......63 Results for the Female Heterosexual Prime Condition............. ................... ...64 Qualitative Results ............................ ...64 DISCUSSION........................ .............. .......66 Aggressive Trait Rating Scale Interaction.........67 Main Effects for Gender......... ........ ..........72 Nurturant Prime Experiment ..................... ...74 Limitations of the Study... .............. .........81 Future Directions for Research .............. ......86 Conclusion .......... ..... ...................... ...88 vi APPENDICES A: Tables and Figures ............................. 91 B: Demographic Questionnaire ..................... 107 C: Priming Stimuli ............................... 108 D: Pilot Study ................................... 123 E: Target Paragraphs ............................. 125 F: Filler Task ................................... 129 G: Sentence Stems ................................ 134 H: Scoring Manual for Sentence Stems ............. 138 I: Trait Rating Scales ........................... 150 J: Consent Forms ................................. 158 K: Feedback Sheet ......... L ...................... 160 REFERENCES ............................................ 161 vii Table Table Table Table Table Table Table Table Table Table Table Table 10 ll 12 LIST OF TABLES Percentage of Women Child Sexual Abuse Perpetrators ........................... 91 Analysis of Variance for Aggressive Trait Rating Scale ........................... 94 Analysis of Variance for the Sentence Completion Measure .................. 96 Analysis of Variance for Nurturant Trait Rating Scale ........................... 97 Analysis of Variance for Male Sexuality Trait Rating Scale ........................... 99 Analysis of Variance for Female Sexuality Trait Rating Scale .......................... 100 Number of Responses to Question 5 Categories ................................ 101 Means for Aggressive Trait Rating Scale ..... 102 Means for Nurturant Trait Rating Scale ...... 103 Means for Male Sexuality Trait Rating Scale ................................ 104 Means for Female Sexuality Trait Rating Scale ................................ 105 Means for the Sentence Completion Measure...106 viii LIST OF FIGURES Figure 1 Aggressive Prime Interaction............95 Figure 2 Nurturant Prime Interaction.............98 ix INTRODUCTION Child sexual abuse, a long denied problem, has received increased attention over the last 15 years: during the 19808 there was an explosion in the scientific and popular literature regarding the sexual abuse of children (Allen, 1991). Victims and perpetrators of this crime became the focus of research and study, but the vast majority of this literature focused on female victims and male perpetrators. (Lew, 1988). In the late 19805 there was an increased focus on male victims of childhood sexual abuse (Finkelhor, 1990). However, to date little scientific research explores the possibility of females perpetrating this type of sexual crime (Mathews, Matthews & Speltz, 1989). A review of the incidence and prevalence data, as well as theories of the etiology of child sexual molesters, demonstrates that women are generally absent from this research (Mayer, 1992). The absence of women in the assessment measures developed for child molesters, the classification scales used, and the theories developed to explain the behavior imply that women do not sexually abuse children. Yet evidence does exist that women sexually abuse children in greater numbers than is usually assumed (Finkelhor & Russell, 1984). 2 It has been suggested that any statistics regarding the number of female perpetrators is inaccurate for a number of reasons (Banning, 1989; Groth, 1979a; Plummer, 1981). Cultural taboos about women's sexuality interfere with the unbiased study of deviant female sexual behavior, and may serve to decrease the number of reported female molesters (Hunter, 1990). Researchers studying female sexual offenders have argued that sexually abusive behavior is "masked" as normal child care behavior when committed by women, and hence, not perceived or reported (Banning, 1989). In addition to the cultural taboos about female sexuality and the effects of masking, gender stereotypes could also affect the perceptions that observers have of female child molesters. Limited experimental data indicates that men and women are indeed judged differently when they sexually abuse a child (Broussard, Wagner & Kazelskis, 1991; Finkelhor & Redfield, 1984); however, to date no study has experimentally explored why this might be so. Research on implicit stereotyping offers a theoretical explanation for understanding why the same behavior could be perceived differently when performed by men versus women. Implicit stereotyping posits that individuals have unconscious stereotypes which are easily evoked and used to judge others without the individual's awareness (Banaji & Greenwald, 1994). Research using this theory has demonstrated the existence of unconscious gender stereotypes 3 (Banaji, Hardin & Rothman, 1993). Ambiguous behavior performed by men is viewed as more aggressive, whereas ambiguous behavior performed by women is perceived as more dependent because of widely held cultural stereotypes of men and women (Banaji et al., 1993). This study will explore the effects of unconscious gender stereotypes, specifically aggression, sexuality, and nurturance, on participant's perceptions of a man or a woman in a sexually ambiguous situation with a child. In addition, to evaluate the possibility that women's sexual activity with children is masked as normal child care, participants' explanations for the ambiguous behavior will be assessed. Childhood Sexual Abuse Sexual abuse of children is a crime that both lay persons and mental health professionals have long been reluctant to acknowledge (Herman & Hirschman, 1981). As recently as 1975, one type of sexual abuse, father-daughter incest, was still cited in a major psychiatric reference text as a crime that occurred in only one case per million (Henderson, 1975). More recent statistics demonstrate that the sexual abuse of children is not such a rare event. For example, a government incidence study indicates that in 1992 alone there were approximately 130,000 cases of substantiated child sexual abuse in the United States (U.S. Department of Health and Human Services, 1994). 4 Conservative prevalence estimates indicate that 27% of women and 16% of men experience some type of direct physical sexual contact with an adult before the age of eighteen (Finkelhor, Hotaling, Lewis, & Smith, 1990). These recent statistics confirm that sexual abuse is a childhood experience for many children. The growing awareness of this problem has resulted in a dramatic increase in the psychological and psychiatric research on this topic (Allen, 1991). A search of a psychology journal citation database revealed that from 1980 to 1990 there were 860 published papers on child sexual abuse; in the 10 years prior, only 38 papers had been published on this topic. The research on childhood sexual abuse has encompassed a wide variety of topics including: 1) understanding and describing the short and long term effects of childhood sexual abuse on victims; 2) descriptions of the types of adults who perpetrate this crime; and 3) the development of theories to explain why some adults find children sexually stimulating and act on these feelings. Research from these three areas of study will be reviewed. Effects of Childhood Sexual Abuse on Victims Numerous studies indicate that childhood sexual abuse causes short and long term psychological damage to many of its victims (Browne & Finkelhor, 1986; Burgess & Hartman, 1987; Finkelhor, 1990; Gale, Thompson, Moran & Sack, 1988; Hunter, 1991; Johnson & Shrier, 1987; Kendall-Tackett, 5 Williams & Finkelhor, 1993; Naitove, 1988; Tharinger, 1990). Particularly damaged are those children who were abused by a relative, molested for a long period of time, subject to repeated incidents, or who were orally, anally, or vaginally penetrated (Kendall—Tackett et al., 1993). Initially a child who has been sexually abused may experience school problems, excessive fears, and severe depression (Saslawsky & Wurtele, 1986). Browne and Finkelhor (1986) cite reactions of anger, hostility, and anxiety, as well as inappropriate sexual behavior in children who have been sexually molested. Immediately after the abuse, one-fifth to two-fifths of victims manifest clinically pathological disturbances (Browne & Finkelhor, 1986). Although one-third of children have no observed or reported symptoms (Kendall-Tackett et al., 1993), the most recent research indicates that girls who are sexually abused may experience negative hormonal, pubertal, and neuroendocrine changes compared to controls (DeAngelis, 1995). The problems for victims of child sexual abuse do not end with childhood. Adults who were abused as children report marital, familial, and interpersonal relationship problems (Saslawsky & Wurtele, 1986), including difficulties parenting their own children (Cole & Woolger, 1989). Other possible long-term effects include depression, self- destructive behavior, anxiety, feelings of isolation and 6 stigma, a tendency toward revictimization, and problems with substance abuse (Browne & Finkelhor, 1986). Not every victim experiences extreme long-term symptoms; when compared to their nonvictimized counterparts, less than one-fifth of adult survivors manifest clinically pathological disturbances. Thus, while extreme long-term effects are not inevitable, the potential for initial and/or permanent mental health damage is a significant problem for the victims of childhood sexual abuse (Browne & Finkelhor, 1986; Finkelhor, 1990). Perpetrators of Child Sexual Abuse Demographic data has been collected on male sexual offenders, including pedophilesl,since the 19305 (Gillespie, 1935 cited in Bender & Blau, 1937). These reports typically list the following information about the offenders: type of sexual crime committed (including incest and pedophilia), age, race, national origin, religion, education, marital status, and previous offenses (e.g., Apfelberg, Sugar & Pfeffer, 1944; Erickson, Walbek & Seely, 1987; Gebhard, Gagnon, Pomeroy & Christenson, 1965; Groth, 1979a). Autobiographical details and specific sexual history information has been collected from men who have sexually abused children (e.g., Erickson et al., 1987). 1The term pedophile as used in this dissertation does not necessarily meet DSM-IV diagnostic criteria; rather, the term is simply used to label those adults who have sexual relations with children. 7 These surveys have provided useful information that challenge common myths about child molesters. For example, the typical child sexual abuser is not a "dirty old man" who is a stranger to the child; rather, in approximately 86% of the reported cases, perpetrators are either related to or well known to their child victims (Cupoli & Sewell, 1988). In addition to demographic and statistical data, unique assessment methods and classification taxonomies have been created for use with child molesters. These evaluation measures collect more specific information about pedophiles. Phallometric assessment or penile plethysmography has been used as a method to measure a man's sexual preference for immature physiques (Freund & Blanchard, 1989; Haywood, Grossman & Cavanaugh, 1990). A device placed around the penis measures penile arousal response to a variety of erotic and neutral slides or film strips, including pictures of adults and children clothed and/or nude (Freund & Watson, 1991). Few men voluntarily admit to a sexual preference for children; thus, these assessment devices aid in making the pedophile diagnosis (Freund & Watson, 1991). As well as physical assessment devices, scales measuring specific psychological functions have been developed to assess men who sexually abuse children. Such a scale is the Abel & Becker Cognitigngistortion Scale (Abel, Gore, Holland, Camp, Becker, & Rathner, 1989). The use of cognitive distortions and denial has been suggested as a 8 defense mechanism typically used by the male pedophile (French, 1990). This scale has been factor analyzed using child molesters, non-child molesting paraphiliacs and a non- paraphiliac control group and can successfully differentiate between child molesters and non-child molesters (Abel et al., 1989). Although the scale can be administered to either men or women, it was developed using only male participants, and published research using the scale has only been with men (J. V. Becker, personal communication, March 8, 1995). There are a number of classification taxonomies for child sexual abusers and they too are used only for males. For example, one of the oldest and most accepted taxonomies is Groth's (1978) differentiation between fixated and regressed offenders (Simon, Sales, Kaszniak, & Kahn, 1992). Groth (1978; 1982) distinguishes between offenders whose primary sexual interest is and always has been children (fixated) and those who have a history of age-mate relationships but, due to severe life stress, regress and begin to have sexual relations with children. In this classification system, fixated offenders are considered "true" pedophiles because, due to their arrested psychosexual development, they are exclusively attracted to children, whereas regressed offenders sexually abuse children only when life demands exceed their abilities to cope. 9 These differentiations have important recidivism and treatment implications; fixated perpetrators have a very poor treatment prognosis, whereas regressed offenders are more responsive to intervention (Barnard, Fuller, Robbins & Shaw, 1989; Finkelhor, 1986; Giarretto, 1987). Because Groth did not include women offenders when he devised this classification taxonomy, it is impossible to know if the dichotomy and its recidivism and treatment implications are applicable to women pedophiles. Etiological Theories of Pedophilia A wide variety of theories have been proposed to explain why some adults find children sexually arousing and act on these feelings (Araji & Finkelhor, 1986). Once again, as with the assessment techniques and classification taxonomies discussed above, etiological theories have been developed only to explain the behavior of male pedophiles: these theories are generally developed in such a manner as to preclude the existence of women who sexually abuse children. Three prominent theories will be discussed, but many more have been proposed (for reviews see Araji & Finkelhor, 1986 and Howells, 1981). Gender socialization patterns, what are typically called feminist explanations of sexual abuse (Araji & Finkelhor, 1986), have been implicated in producing males who sexually abuse women and children (Allen, 1991). This theory posits that males are socialized to value dominance 10 and aggression, to initiate sexual relations, and to desire sexual partners who are youthful and subservient (Araji & Finkelhor, 1986). Sexual abuse of women and children is a natural outgrowth of these values (Allen, 1991). Male pedophiles, who feel uncomfortable and shy around adult women, are sexual with children because children are weak and nonthreatening sexual objects that these men can dominate (Howells, 1981). According to this theory, women are socialized to seek older more powerful mates, and not to initiate sexual relations; thus, women would not sexually abuse children (Allen, 1991; Howells, 1981). This theory provides no explanation for women who sexually abuse children (Allen, 1991). Psychoanalytic explanations for pedophilia result in a similar problem: it is impossible to explain why a woman would sexually abuse a child based on psychoanalytic principles. Pedophiles are believed to have intense and unresolved unconscious conflicts with their mothers concerning separation/individuation that prevent them from successfully working through the oedipal stage of development (Howells, 1981; Socarides, 1991). Severe castration anxieties result from these conflicts and prevent pedophiles from developing adult social and heterosexual behavior; they are unable to relinquish their relationship to their mothers so they might identify primarily with their fathers (Araji & Finkelhor, 1986; Howells, 1981; Socarides, 11 1991). Children, for lack of anything better, are chosen as a substitute for adult sexual relations (Araji & Finkelhor, 1986). Sex with children can also serve as a defense against the more psychologically threatening relationship with an adult woman. As with the gender socialization theory, psychoanalytic explanations cannot explain female pedophilia. This theory is rooted in a man's conflicted relationship with his mother and his inability to identify with his father and thus develop normal adult heterosexual relationships. No analogous theory has been proposed to explain how abnormal female development could result in a pedophilic sexual perversion. Unlike the two theories described above, biological theories are not constructed to exclude women, they are just primarily applied to men. The premise of biological theories is that physical or chemical abnormalities compel an otherwise normal individual to act in a sexually deviant manner (Langevin, 1992). Brain damage of an electrical, structural, functional, and chemical nature has been associated with paraphiliac behavior. Langevin (1992) has examined computer tomography (CT) scans and found that pedophiles tend to show dilation of the left temporal and anterior horns; in fact, they show more brain damage (less dense right frontal-temporal areas, more asymmetry, smaller frontal and temporal areas) than any 12 other paraphiliac group. Hendricks, Fitzpatrick, Hartmann, Quaife, Stratbucker, and Graber (1988) reached a similar conclusion, also using CT scans. They found that child molesters had thinner and less dense skulls and lower regional cerebral blood flow compared to controls. Why these specific brain anomalies would lead to paraphiliac activities has not been explained. Chemical abnormalities within the endocrine system have also been found in sexual deviants involving the levels of luteinizing hormone, follicle stimulating hormone, and testosterone (Langevin, 1992). Pedophiles over-secrete luteinizing hormone. Elevated testosterone levels have been associated with sexually violent behavior (Langevin, 1992). Some reports of elevated testosterone in pedophiles exist, but other researchers have failed to replicate these findings (Araji & Finkelhor, 1986). A single case study of a 20-year-old female pedophile using extensive assessments (clinical, psychometric, endocrinologic, electroencephalographic, and vaginal photoplethysmographic) found no evidence of endocrine abnormalities (Cooper, Swaminath, Baxter, & Poulin, 1990). Cooper et al. (1990) recommended using such a battery when evaluating women suspected of child sexual abuse, but no other examples of this type of rigorous evaluation of female offenders has been found in the literature. The biological theories to explain pedophilia might be applicable to women 13 sexual molesters if the appropriate data were gathered. Information about child victims and their sexual abusers has increased greatly since the 19703; however, this research has only expanded knowledge of the male perpetrator. The absence of female perpetrators from the literature on child sexual abuse raises many questions. It is obviously not possible to publish demographic data about female pedophiles nor necessary to include women in the development of assessment scales, classification taxonomies, and etiological theories if women do not sexually abuse children. If, on the other hand, women do sexually abuse children, one cannot assume that such women are demographically similar to the men who abuse children, or that women use the same defense mechanisms as male abusers, or that women would fit into a classification taxonomy that has been designed specifically for men, or, finally, that theories of etiology created to explain male behavior would apply to women. The general lack of reference to female child sexual abusers in the published literature seems to imply that women do not sexually abuse children. There has, however, been some published reports on female pedophiles, especially since 1985 (Allen, 1991). The percentage of child sexual abusers who are women varies widely between studies; statistics cited range from .2% to 59% (see Table 1 in Appendix A). 14 Percentage of Female Child Molesters. The statistics most commonly cited, those of incarcerated pedophiles, would lead one to conclude that under 2% of child molesters are women (e.g., Musk & Gallagher, 1985 report 1.65%; O'Connor, 1987 reports .2%; Rowan, Rowan & Langelier, 1990 report 1.5%). These data accurately reflect the number of women pedophiles in prison, but these figures probably underrepresent the actual number of women child sexual abusers for a number of reasons. Incarceration represents the end point in a long chain of events. First a report of child sexual abuse has to be made, then an individual has to be charged with the crime. Next a decision is made as to whether the individual should be brought to trial. Then the person is tried, and if found guilty, a sentencing decision is made. Finally, the perpetrator has to be incarcerated. Female offenders drop out at each point in this chain of events. Observers do not expect women to sexually abuse children and may not label suspect behavior as sexually abusive (Banning, 1989). Because children have more difficulty reporting sexual abuse that involves women perpetrators, compared to abuse that involves men (Kendall-Tackett & Simon, 1987), women are probably charged less often than are men. Women are typically not adjudicated for sexual crimes (Groth, 1979a; Williams & Farrell, 1990), and only the most severe sexually abusive acts committed by women are reported (Finkelhor, 15 Williams & Burns, 1988; Travin, Cullen, & Protter, 1990). Finally, when adjudicated, women are often given treatment sentences, not incarceration sentences (Allen, 1991; Gebhard et al., 1965). Interestingly, the highest reported figures for female sexual offenders also come from prison data. Male inmates convicted of sexual assault, who were sexually abused in childhood, report that from 41% (Groth, 1979b) to 59% (Petrovich & Templer, 1984) of their abusers were women (see Table 1). This is a surprisingly high percentage of female perpetrators. It has been suggested that either childhood sexual assault by a female plays a causative role in the formation of some male sexual abusers (Freeman-Longo, 1986) or that these men are lying about a history of sexual abuse (Wakefield & Underwager, 1991). Data from this sample cannot be generalized to the population as a whole, as incarcerated male sexual offenders are clearly a very unrepresentative sample of the general population. It is likely that the actual incidence of female sexual offenders falls somewhere between these high and low figures. The most accurate figures may be those provided by Finkelhor and Russell (1984) who re-analyzed the 1981 National Incidence Study of Child Abuse and Neglect (N18) and the 1981 American Humane Association (AHA) study of child sexual abuse. These two data sets reflect data gathered from a wide variety of reporting agencies and 16 sources. Cases in the National Incidence Study were reported by professionals in 26 counties chosen to be representative of the entire United States. Cases were reported to the NIS even if they were not reported to official agencies. Data from the American Humane Association study were reported to a mandated reporting agency in 31 states. Finkelhor and Russell's sample is probably more representative of the population in the United States than the other studies cited in Table 1. Thus, their conclusions are the most accurate estimate, to date, of the percentage of sexual offenders who are women. I Both studies initially reported a very high percentage of female offenders (37% in the NIS report and 46% in the AHA report) because women were included as sexual abuse offenders if they permitted sexual abuse to occur to their children (Finkelhor & Russell, 1984). Thus, these high figures included a sizable percentage of women who did not themselves physically participate in sexual activities with children. When Finkelhor and Russell (1984) reanalyzed the data and omitted this group of women, the percentages of female offenders were revised downward: 24% of boys and 13% of girls in the NIS study and 14% of boys and 6% of girls in the AHA study were abused by women. These statistics also included some women who abused in concert with a man, and may have been coerced into participating. After further analyses to try to isolate females acting alone, not in 17 conjunction with a male, Finkelhor and Russell concluded that the best estimate of female offenders of child sexual abuse is 20% (range 14 to 27%) for male victims and 5% (range 0 to 10%) for female victims. Unfortunately, most recent government reports on child sexual abuse do not routinely include gender of the perpetrator (see National Center on Child Abuse and Neglect, 1992 or 1993). An exception is data from the U.S. Department of Health and Human Services (1994) which reports that 18% of the child sexual abuse perpetrators in North Dakota for 1992 were women. This figure was not broken down by gender of the victim, but it generally supports the Finkelhor and Russell conclusion from the NIS and AHA 1981 data. Still, the Finkelhor and Russell statistics on percentage of female sexual offenders is an estimate and subject to a host of problems that could render it inaccurate. Statistics on incidence and prevalence of child sexual abuse (whether committed by a man or a woman) are of dubious reliability for numerous reasons, including the lack of a standard definition for sexual abuse and the use of very different subject pools and methodologies (for reviews see Haugaard & Emery, 1989; Peters, Wyatt, & Finkelhor, 1986; Wyatt & Peters, 1986a and 1986b). But for female sexual abusers, cultural taboos concerning female sexuality and aggression could serve to reduce further the number of 18 reported female pedophiles. Female Sexuality There still exists a subtle social taboo that prevents the acknowledgement and study of female sexuality, including the study of normal sexuality such as masturbation (Chilman, 1995 quoted in Sleek, 1995; Bite, 1993). The source of this bias is complex and beyond the scope of this paper but it is expressed in popular culture, language, as well as scientific writings and research. Normal Female Sexuality In popular culture (e.g., film, novels), women's sexuality is positively depicted only within the bounds of marriage (Ellis, 1990), and different standards still exist for men and women. The virgin/whore stereotype is alive and well: women who have sex with too many men are labeled sluts, tramps, whores, or nymphomaniacs, whereas a man is seldom given derogatory labels for "sowing his oats" (Kaplan, 1990). Our language has few words to label female sexual experiences but many expressions exist to describe male experiences; there is no vernacular term for "manual clitoral stimulation" but "manual penile stimulation" is called "jerking off," "beating your meat," "choking the chicken," or "spanking the monkey." Female masturbation is still such a taboo subject that there are no colloquial expressions for the action, yet male masturbation is described in verbally colorful and somewhat violent 19 terminology. If we were comfortable with female sexuality there would be analogous terms to describe female masturbation because most women (82%) do masturbate (Hite, 1993). Different words are used to describe the same process in men and women's bodies: male sexual tissue becomes "erect" with blood but female sexual tissue becomes "congested" with blood as though it were a wound. The word erect seems to be too forceful or aggressive to describe a woman's body, yet the exact same biological process happens to the sexual tissues in men and women (Herschberger, 1948; Bite, 1993). I Female Sexual Perversions Western culture seems much less comfortable thinking about female sexuality than male sexuality, especially sexual perversions (Welldon, 1991). As discussed earlier, this bias is clear in the literature on sexual perversions. Women are not included in both the classifying of paraphilias (Abel, Rouleau & Cunningham-Rathner, 1986) and in the understanding of the genesis of this behavior (Welldon, 1988; 1991). For example, when classifying and describing paraphilias, a modern medical school text notes that women with these sorts of behaviors are "treated as a medical curiosity" (Abel et al., 1986, p. 290). Females with sexual perversions are often ignored or are not judged with the same seriousness as are males (Gebhard et al., 1965). When women paraphiliacs are acknowledged, the 20 behavior is often trivialized or made light of (Mathis, 1972). Individuals, both professionals and the public, often disregard the impact that being sexually abused by a woman has on her child victims (Barnard et al., 1989; Broussard, Wagner & Kazelskis, 1991; Finkelhor & Redfield, 1984). Finkelhor and Redfield (1984) found that participants consistently rated sexual abuse vignettes with a female perpetrator as less abusive than the same vignette with a male perpetrator. Broussard et a1. (1991) found that female perpetrators of male victims were rated as significantly ' less representative of child sexual abuse than the three other gender combinations, and as less harmful to the victim. This belief, that women are sexually harmless, was most clearly revealed by Mathis (1972) in a chapter on pedophilia. He states that "... the female is conceptualized as being sexually harmless... That she might seduce a helpless child into sexplay is unthinkable, and even if she did so, what harm can be done without a penis?" (p. 54). When a child is sexually abused a relationship of trust between an adult and a child has been violated. Damage from sexual abuse is usually psychological and emotional, not physically harmful (Meiselman, 1978), regardless of the gender of the perpetrator. The presence or absence of a penis, in and of itself, should not be used as the sole criterion to determine the harmfulness of sexual abuse. 21 Many more men than women are diagnosed with paraphilias, including pedophilia, but the huge discrepancy that is cited could, in part, be due to an ignoring or denying of female perversions (Welldon, 1991). It is difficult for our culture to discuss and label even normal female sexuality except in certain situations (i.e., marriage): this subtle ignoring of women's sexuality could also influence both the recognition and labeling of women with sexual perversions. In other words, cultural biases that inhibit the studying of normal female sexuality might also affect the study of sexual perversions and serve to reduce the number of women who are defined as paraphiliacs, including pedophiles. Adult Female Sexual Contact with Boys A second curious effect of the prohibitions surrounding women's sexuality is that adult female sexual activity is also tacitly ignored in certain situations. There is a subtle cultural norm that labels sexual activity between boys and women as normal but sexual relations between girls and men as sexual abuse. The myth that males cannot achieve or maintain an erection when threatened (Sarrel & Masters, 1982) serves to normalize this sexual activity further. After all, we reason that a boy would not respond sexually if he did not want to. Adult males who have disclosed sexual abuse to family and friends are told that they were "lucky" to have had a 22 sexual experience at such a young age, independent of how violated or exploited they feel (Hunter, 1990). This stereotype is reinforced in the media, especially in films (e.g., Murmurs of the Heart) which depict sexual experiences between boys and women as a rite of passage. If boys do not enjoy such experiences, their masculinity or sexual identity is called into question (Hunter, 1990). Certainly some adolescent males have consensual sexual relationships with older women, but when women in positions of authority use that power to initiate sexual contacts with boys, the women have abused their power and they have sexually abused a child. Masked Female Sexual Contacts Numerous authors writing about female sexual abuse have argued that sexually abusive behavior is masked when committed by women (e.g., Banning, 1989; Groth, 1979a: Justice & Justice, 1979; Kempe & Kempe, 1984; Plummer, 1981). Masking refers to the disguised nature of female sexual abuse; women may be able to touch children inappropriately and it is viewed as maternal, but men who touch children in a similar manner might be looked upon with suspicion (Banning, 1989; Goodwin & DiVasto, 1979; Groth, 1979a; Plummer, 1981). The above writers suspect that there is a great deal of adult female-child sexual behavior or contact that goes unreported because it is disguised as "normal" maternal child care. 23 It is plausible that women are able to disguise sexual activity with children through caretaking activities; they have much greater access to children's bodies because of diapering, bathing, and dressing them than men usually do. Women also generally administer health care needs to children, such as enemas. Because of the greater access and expected intimate contact between women and children, sexual abuse by a mother or caretaker could be either hidden by the woman or misperceived by observers as normal caretaking activities. Finkelhor et al. (1988), in their study of sexual abuse in daycare centers, found that women were only reported for severe acts of sexual abuse, whereas for male perpetrators, a wide range of sexual activities were reported. Perhaps the less abusive activities were misperceived when perpetrated by women. Cultural taboos surrounding female sexuality that are expressed in both popular culture, as well as scientific writings, could reduce the number of reported female pedophiles in three ways: by ignoring women who engage in overt sexual perversions, by permitting women to have sexual contact with underage boys, and by misperceiving sexual contact between female caretakers and children as normal child care activities. Another taboo that could affect perceptions of women who sexually abuse children are culturally learned unconscious stereotypes concerning women and aggression: 24 women are not expected to be aggressive. These implicit stereotypes could make it difficult for observers to recognize that what they were witnessing was sexual abuse, especially if it was committed in the context of caretaking activities. Such stereotypes, in conjunction with the above cultural taboos surrounding female sexuality, could decrease the number of women reported for child sexual abuse crimes by making sexual activities between women and children much more difficult for observers to recognize than such activity between men and children. Implicit (Unconscious) Gender Stereotypigg Gender Stereotypes Stereotypes of men are generally more positive than stereotypes of women (Broverman, Vogel, Broverman, Clarkson & Rosenkrantz, 1972). Deaux argues that stereotypes generally favor men over women, and different explanations are suggested for the same behavior depending on the gender of the actor (Deaux, 1976 cited in Fiske & Taylor, 1984). For example, successful men are perceived as intelligent (an internal, stable quality) but successful women are perceived as hard workers or lucky (unstable factors) (Fiske & Taylor, 1984). Gender stereotypes can also serve as a protective mechanism for women. Employment gender bias against women is less recognized when it comes from other women because participants expect such stereotyping from males, not 25 females (Baron, Burgess & Kao, 1991). Farrell (1993) cites numerous statistics showing that for the exact same criminal offense, women, as compared to men, receive significantly shorter sentences, treatment sentences rather than incarceration, and lower bail. Farrell (1993) argues that this is because women are seen as in need of protection, innocent of criminal behavior, and are not held responsible for their actions when they do commit a crime. It has been demonstrated that harsher punishments are recommended for behavioral transgressions when the transgressions are consistent with an ethnic stereotype (Bodenhausen & Wyer,l 1985); it is possible that this is true for gender stereotypes and may contribute to women receiving different criminal punishments than men. Perceptions of stereotypical masculine and feminine personality traits have not changed significantly in the United States since they were first studied in the early 1970s (Bergen & Williams, 1991). Traditional gender stereotypes describe men as adaptive-instrumental or agenic and women as integrative-expressive or communal (Bergen & Williams, 1991; Gerber, 1991). Agenic men are dominant leaders who tell others what to do and they are obeyed (Rothbaum, 1977). Communal women are nurturant and kind to others, especially those in need (Rothbaum, 1977). Stereotypes of sexuality follow a similar pattern: male heterosexuality is perceived of as dominant and aggressive, 26 whereas female heterosexuality is seen as passive and receptive (Hite, 1993; Hyde & Rosenberg, 1976). Women are stereotypically perceived as not interested in sex and as not enjoying sexual activities as much as men (Garcia, 1982); whereas men are portrayed as always interested in and ready for sexual activity (Hite, 1993). Women are not stereotypically associated with sexual aggressiveness because of their weaker sex drive; but males are stereotypically associated with sexual aggressiveness (Gebhard et al., 1965; Hite, 1993; Sarrell & Masters, 1982). Thus, men are stereotypically associated with sexual crimes (including child sexual abuse) but women, because of the "passive" nature of their sexuality, are not believed capable of committing such crimes (Gebhard et al., 1965). In the past 20 years there has been a considerable decline in the overt expression of gender stereotypes, yet an insidious form of sexism remains in today's culture (Banaji & Greenwald, 1994). Implicit or unconscious stereotypes affect how a person perceives and judges situations. These stereotypes operate without an individual's awareness that they are using traditional gender stereotypes to evaluate someone (Banaji & Greenwald, 1994). The study of unconscious mental processes, including implicit stereotypes is a relatively new field (Greenwald, 1992). 27 Unconscious Cognitive Processes Recent research on implicit memory and learning has resulted in a re-examination of the role of the unconscious in cognitive processes (Greenwald, 1992). Research has revealed the existence of implicit or unconscious memory in the absence of explicit or conscious recall (Lewicki, Hill & Czyzewaka, 1992; Schacter, 1992). Using indirect measurements of immediate or long-term residues, the influence of material that was barely perceptible, perceptible but not attended to, or attended to but forgotten is apparent (Greenwald, 1992). That is, participants cannot consciously recall this material, but through indirect measurement it is apparent that participants are influenced by unconscious learning (Banaji & Greenwald, 1994). This type of unconscious cognition is generally believed to be intellectually simple compared to the psychoanalytic construct of the unconscious2 2There appear to be at least three definitions of the term unconscious currently in use. In the psychoanalytic literature the term refers to that part of the brain that holds or stores repressed knowledge and is able to employ complex defense mechanisms to prevent an individual from experiencing overwhelming anxiety connected to the repressed information. In the study of cognitive psychology the term is used to describe that part of the brain that controls nonconscious information acquisition processes. Sometimes this part of the brain is described as faster and structurally more sophisticated than consciously controlled cognition (Lewicki, Hill & Czyzewska, 1992; Erdelyi, 1992). But other cognitive psychology researchers describe it as very limited in its analytic abilities (Bruner, 1992; Greenwald, 1992). Finally, in social psychology research on stereotypes, the term unconscious or implicit is used to describe the automatic application of sociocultural beliefs 28 (Greenwald, 1992). Implicit (Unconscious) Stereotypinq Research about and definitions of stereotypes have generally neglected the influence of unconscious mental processes (Banaji & Greenwald, 1994). Traditional theory and methodology used to study stereotypes precluded the existence of unconscious stereotypes by relying on direct, self-report measures. This methodology, and by association the underlying theories, assumes the conscious operation of stereotypes, and therefore cannot be used to measure and study implicit stereotypes (Banaji & Greenwald, 1994). Implicit stereotyping is the unconscious application of information about an existing relationship between an attribute (e.g., dependence or aggression) and a social category (e.g., race or gender). It is the act by which an individual replicates beliefs held by society as a whole: common knowledge or beliefs about a group is unconsciously applied when judging an individual from that group (Banaji & Greenwald, 1994). People who share a common culture also share common implicit stereotypes. Individuals possess both cultural stereotypes and personal beliefs about individuals from specific groups; there can be overlap between these two sets of beliefs that have been internalized by a subject without the subject's awareness and are used unconsciously to judge target individuals. 29 (Devine, 1989). Children internalize cultural stereotypes long before they are able to evaluate the appropriateness of the stereotypes and create their own personal beliefs (Devine, 1989). Within an individual, these cultural stereotypes have a long history of activation. Because they are internalized at an early age, they are a well-learned set of associations, and they are automatically activated when a participant is in the presence of an individual from a target group or the symbolic equivalent (Devine, 1989). This process is both automatic (Devine, 1989) and unconscious (Banaji & Greenwald, 1994). People who share a similar cultural upbringing also possess similar unconscious stereotypes, even if their conscious beliefs are very different. The unintentional or unconscious activation of cultural stereotypes is equally strong for participants whose personal beliefs mirror those of the cultural stereotypes and for those whose personal beliefs do not (Devine, 1989). Because cultural stereotypes have a longer history of association and activation within an individual's cognitive processes than do personal beliefs, the development of personal beliefs can create a conflict between the already well-established cultural stereotypes and the new beliefs (Devine, 1989). A person who has consciously rejected an implicit stereotype will still find it difficult to inhibit implicit stereotyping when the individual is not focusing on the stereotype in 30 question (Devine, 1989). This has been found to be true specifically for gender stereotypes; when gender is not salient, participants judge themselves and others with traditional gender traits (Higgins & King, 1981 cited in Devine, 1989). Implicit (Unconscious) Gender Stereotypes A number of recent studies have demonstrated the operation of unconscious gender stereotyping (Banaji et al., 1993; Banaji & Greenwald, 1995; Blair & Banaji, under review). In these three experiments, participants' implicit stereotypes are activated using a priming task. The purpose of priming is to stimulate or activate unconscious stereotypes (e.g., gender or racial stereotypes) much as would happen in every day life. For example, an individual's unconscious stereotypes would be activated while watching a television news program in which a number of African American males are implicated in a number of violent crimes. This individual's stereotype that African American males are aggressive and violent has been activated. Later, in an unrelated situation, this same individual unconsciously applies this stereotype and judges an ambiguous behavior performed by an African American male as aggressive. In the laboratory, using behavioral examples of the stereotype of interest serves to activate a participant's unconscious stereotypical schemas (Srull & Wyer, 1979). The 31 more behavioral examples used, the greater the effect of priming (Srull & Wyer, 1979). The participant is unaware that implicit stereotypical associations have been activated. These stereotypical beliefs or schemas could be applied to subsequent unrelated situations and used to mediate judgements, but only if the judged person fits the activated social category (Banaji, Hardin & Rothman, 1993; Greenwald & Banaji, 1995). To make the experiment realistic, participants cannot be aware that they are being primed. This can be done in a number of ways--through the use of subliminal priming (e.g. Devine, 1989) or by presenting the priming task as a separate and unrelated experiment (e.g. Banaji et al., 1993)3. After priming, a manipulation is carried out and then measured. There are three issues concerning the effects of priming that will be briefly reviewed: the length of delay that is needed between the manipulation and the administration of the dependent measure in order to obtain a priming effect, whether or not the primed trait is 3A number of priming designs have been used by social psychologists to study unconscious stereotypes [subliminal exposure (see Devine, 1989), semantic priming (see Blair & Banaji, under review), and the false-fame procedure (see Banaji & Greenwald, 1995)] but these three procedures do not readily lend themselves to a vignette experiment. For example, the simple and rapid fire responses given in a semantic priming experiment would not be applicable to the reading and judgment of characters in a vignette design. For the proposed study, the most appropriate design would be the priming procedure used by Banaji et al. (1993). 32 transferred to other, related schemas, and the effects of chronic environmental exposure on stereotypes when priming. The effects of priming increase with time if there is a delay between reading the vignette and the administration of the dependent measure (Srull & Wyer, 1979). Researchers have had participants rate the target immediately, a day later, or a week later (Srull & Wyer 1980) and found that the effects of priming are greater as the delay increases. It has been suggested that situational information that does not support or contradicts the primed stereotype is forgotten during the delay, making the effect of the primed stereotype stronger (Srull & Wyer, 1980). But other investigators, using only a 10-minute delay between reading the vignette and rating the target, have found priming effects (Banaji et al., 1993). A comparison of Banaji et al.’s (1993) and Srull and Wyer's (1980) mean trait ratings (both used 10—point Likert scales) demonstrated that a longer delay did not necessarily result in higher means for the primed trait. Thus, having participants return to the lab a day or a week later to complete the dependent measure is not necessary for priming to have an effect. The second issue concerns the effects of priming on other related traits. The literature indicates inconsistent findings as to whether priming transfers to other semantically (cognitively) and evaluatively (positively or negatively) congruent traits (Erdley & D'Agostino, 1988). 33 Some studies have found that priming effects transfer only to semantically related trait dimensions (Devine, 1989; Erdley & D'Agostino, 1988). Yet other studies find that the priming effect only generalizes to traits that are evaluatively related but descriptively unrelated to the primed trait (Srull & Wyer, 1980). Banaji et a1. (1993) designed their trait rating scale to test for the possibility of priming extending to other traits which were semantically or evaluatively related to the primed trait. Only the primed stereotype demonstrated a prime type by gender interaction; the other 16 traits on their scale did not. In addition to the effects of laboratory priming, participants in a study are also influenced by chronic exposure to environmental priming. This source of priming could be the same that originally created an individual's stereotypes (e.g., gender stereotyping in a household, the medias' portrayal of certain categories of people, or a school's culture). Chronic exposure to stereotypes leads to the formation of specific constructs that then are easily, or chronically, activated for those individuals who possess the chronic construct (Bargh, Bond, Lombardi, & Tota, 1986). For example, an individual raised in a religious household or environment would have a chronically accessible construct of "spirituality" which would, because of its accessibility for this individual, be used to interpret social and 34 behavior stimuli from the environment. An individual who is not raised in such an environment, and not exposed to such stereotypes, would not be expected to possess a chronically accessible construct of spirituality. When an individual possesses a chronically accessible construct, and participates in an experiment in which that construct is primed for, the chronically accessible construct interacts with the priming and produces a greater effect because there are now two sources of accessibility, the chronic construct, and the temporary priming stimulation (Bargh et al., 1986). To determine if an individual has a chronically accessible construct, it is necessary to measure the specific construct (Bargh et al., 1986). Since the chronic accessibility of the stereotypes that will be primed for in this experiment will not be measured, it will not be possible to predict who has chronically accessible stereotypes but for those individuals who do, the priming will have a more powerful effect. Unconscious stereotypes cannot be measured using traditional self-report measures (Greenwald & Banaji, 1995): therefore, it is necessary to use indirect measurement techniques when studying unconscious stereotypes. "Indirect measures are identifiable chiefly . . . by their not alerting the participant to the identity of the [stereotype] being measured" (Greenwald & Banaji, 1995, page 12). The type of dependent measure typically used to assess the 35 effects of priming are trait rating scales (cf. Banaji et al., 1993; Erdley & D'Agostino, 1988; Srull & Wyer, 1979, 1980). Higher ratings on the dependent measure are predicted on the stereotyped trait for individual's whose social category (i.e., gender) is associated with the previously primed stereotype. This is referred to as an assimilation effect (Banaji et al., 1993). But when the primed stereotype is not associated with the target's gender, a number of outcomes are possible: a contrast effect, a weak assimilation effect, or no effect (Banaji et al., 1993). A contrast effect is obtained when primes are negatively associated with a target's social category (Banaji et al., 1993). This results in lower trait ratings because the primed stereotype "act[s] as a standard with which a stereotype-inconsistent target is compared and contrasted" (Banaji et al. 1993, p. 276; also see Srull & Wyer, 1980). For example, Banaji et al. (1993) initially obtained a contrast effect for the male target when participants were primed with a dependent stereotype. Males were rated lower in the dependent condition than the neutral condition because the male target's behavior was judged as less dependent after dependent stereotypes had been activated. Typically one would not expect a change from the dependent to the neutral condition when the stereotype in question does not apply to the target. Contrast effects 36 have generally not been found except with nonverbal stimuli (e.g., photographs judged for quality; Srull & Wyer, 1980). Weak assimilation effects are possible if the relationship between the prime and the target's social category is related to the strength of applicability (Banaji et al., 1993). For example, assertiveness is weakly related to women; one might expect a weak assimilation effect with a female target when assertiveness has been primed for, but not as great as the effect achieved with a male target. Generally no assimilation effect has been found when primes did not apply to a target's social category (Banaji et al., 1993; Erdley & D'Agostino, 1988; Higgins et al., 1977; Srull & Wyer, 1979, 1980). Using methodology similar to that just described, three studies that have found support for implicit gender stereotyping will be briefly discussed. Banaji and Greenwald (1995) have demonstrated stereotypical gender bias in false-fame judgments. Non- famous male names were significantly more likely than non- famous female names to be labeled as famous because individuals were more likely to associate male names with achievement and, thus, famousness (Banaji & Greenwald, 1995). Blair and Banaji (under review) examined the effect that motivation and cognitive resources had on the operation of implicit gender stereotyping. They found that it was very difficult for participants to avoid gender stereotyping 37 even when they were told to do so, unless participants had both the motivation and the resources to counteract automatic stereotyping (Blair & Banaji, under review). Finally, Banaji et al. (1993), using two well- established gender stereotypes (aggressiveness and dependence), demonstrated that participants will unconsciously rate a male target, but not a female target, as more aggressive after exposure to aggression primes, but not neutral primes; a female, but not a male, as more dependent after exposure to dependent primes, but not neutral primes (Banaji et al., 1993). Implicit Gender Stereotypes in an Ambigppus Sexpally Abusive Vignette. Researchers who study female sexual offenders have proposed that sexually abusive behavior is perceived differently when performed by men versus women. It has been hypothesized that when performed by women this behavior is seen as normal "maternal" child care behavior or normal "maternal" affection, but when engaged in by men, this activity is perceived of as sexual abuse. Unconscious gender stereotypes might explain why women's sexually abusive behavior is not recognized or is interpreted differently from men’s actions. Unconscious stereotypes that observers hold about men and women could prevent them from recognizing women as perpetrating sexually aggressive behavior. Research is needed to examine the effect of 38 unconscious stereotypes on perceptions of male and female sexual abusers. Rationale of the Study Women perpetrators are strikingly absent from the increased attention and study that victims and perpetrators of childhood sexual abuse have recently received, even though women molest 20% of boy victims and 5% of girl victims (Finkelhor & Russell, 1984). In part, the underreporting of female perpetrators may be a result of differential perceptions of men and women's intentions. Two studies have found that men and women are judged‘ differently when sexually abusing a child. Finkelhor and Redfield (1984) used a single question to explore individuals' attitudes toward sexual abuse. Participants were asked to rate on a 10—point Likert scale how abusive they found a one sentence sexual abuse vignette. Participants consistently rated sexual abuse vignettes with a female perpetrator as less abusive than the same vignette with a male perpetrator. Broussard et al. (1991), using a vignette involving a 15-year-old adolescent, manipulated gender of the perpetrator and victim, as well as resistance level of the victim. They found that female perpetrators of male victims were rated as a significantly less representative pairing of child sexual abuse compared to the other three groupings (i.e. male perpetrators/male victims, male 39 perpetrators/female victims, and female perpetrators/female victims), and as less harmful to the victim. Neither of these studies sought to explain why men and women are perceived differently when molesting children. This study is the first attempt to do so. Recognizing the existence of female child molesters challenges both conscious and unconscious stereotypes about women. The research literature described earlier indicates that women's heterosexuality, especially sexual perversions, is often consciously denied or ignored. The influence of unconscious sexual stereotypes on perceptions of pedophiles has not yet been studied. Unconscious stereotypes of male heterosexuality could influence perceptions of child sexual abuse because male heterosexuality is also associated with aggression; unconscious stereotypes of female heterosexuality would not be expected to effect perceptions of child sexual abuse because female heterosexuality is stereotypically viewed as passive and would not be associated with sexually abusive behavior. Even if an individual does not consciously ascribe to traditional gender stereotypes, implicit stereotypes are easily activated and difficult to counteract unless a participant is aware of what is being measured and consciously activates his/her personal beliefs. It is possible that these implicit gender stereotypes affect judgements made about males and females who sexually abuse 40 children. This project used two separate studies and a variant of Banaji et al.'s (1993) research design. Participants were primed to elicit implicit aggressive or nurturant stereotypes in the first study. In the second study, the effects of unconscious male and female heterosexual stereotypes were measured. Both studies also contained a control group that received neutral primes, and served as the comparison group for both prime groups. This design allowed the study to measure the effects of unconscious aggressive, nurturant, and sexual stereotypes on judgements of male and female targets in a sexually ambiguous situation with a child. Unscrambling nurturant, sexual, or aggressive sentences acted as a prime to activate participants’ unconscious gender stereotypes. After priming, participants read a vignette describing ambiguous sexual contact between an adult and a child, and completed two dependent measures in which judgments were made about the adult in the vignette. Participants received a story with either a male or a female protagonist. The procedures employed used an indirect measure of gender stereotypes by not alerting participants to the gender manipulation. A comparison of the responses made to the male and female protagonists provided a measure of gender stereotyping without the participant's awareness that 41 the variable under study was gender. An additional change made to the Banaji et al. (1993) methodology was the use of a sentence stem measure (prior to filling out the trait rating scale). This open-ended measure sought to evaluate how participants explained the behavior of the adult in the vignette. This design allowed the researcher to examine participants' explanations regarding ambiguous sexual situations: are women perceived as nurturant and nonsexual in a sexually ambiguous situation and are men seen as aggressive and potentially sexually abusive. Hypotheses of Presept Investigation This project investigated the perceptions that participants had of a man or woman in a sexually ambiguous situation with a child. Participants were assigned to one of two studies; within each study were 2 experiments. Each experiment was a 2 X 2 (prime type by target gender) between-subjects design. In each experiment, the target gender of the vignette protagonist was either male or female. In Study 1 the prime type was either aggressive, nurturant, or neutral. In Study 2 the primes were either male heterosexual, female heterosexual, or neutral. The hypotheses were divided into two sections reflecting the two measures that were used. Septiop I: Implici; Gender sperepgypes It was hypothesized that implicit or unconscious gender 42 stereotypes would differentially effect how men and women in a sexually ambiguous encounter with a child were judged and evaluated. Hypothesis for Study 1. Participants who were primed with aggressive primes will judge male targets as more aggressive than female targets. Participants who were primed with neutral primes will demonstrate no difference when judging male and female targets as aggressive. That is, an interaction between prime type and target gender was predicted for the trait aggressive. After unconscious aggressive stereotypes were activated through priming, it was expected that male targets in an ambiguous vignette would be evaluated as more aggressive than female targets. No difference in aggressiveness rating between male and female targets was expected after exposure to neutral primes. No interactions were predicted between prime type (aggressive or neutral) and target gender for the other 16 traits on the measure (semantically related and unrelated to aggressive). II. Participants who were primed with nurturant primes will judge female targets as more nurturant than male targets. Participants who were primed with neutral primes will show no difference when judging male and 43 female targets as nurturant. In other words, those participants whose implicit nurturant stereotypes have been activated will evidence an interaction between prime type and target gender. After unconscious nurturant stereotypes have been activated, female targets in the sexually ambiguous vignette will be evaluated as more nurturant than male targets. No rating difference in nurturant was expected for male and female targets after exposure to neutral primes. No interactions between prime type (nurturant or neutral) and target gender were predicted for the other 16 traits on the measure (semantically related and unrelated to nurturant). Hypothesis for Study 2. III. Participants who were primed with male sexuality primes will judge male targets as more sexual than female targets. Participants who were primed with neutral primes will demonstrate no difference when judging male and female targets as sexual. Thus, an interaction between prime type and target gender was predicted for the trait male sexuality. After unconscious male heterosexual stereotypes have been activated through priming, male targets in an ambiguous vignette will be evaluated as more sexual than female targets. No difference in sexuality rating between male and 44 female targets was expected after exposure to neutral primes. No interactions were predicted between prime type (male heterosexual or neutral) and target gender for the other 16 traits on the measure (semantically related and unrelated to male heterosexual). IV. Participants who were primed with female heterosexual primes will not judge female targets as more sexual than male targets. Participants who were primed with neutral primes will show no difference when judging male and female targets as sexual. That is, for participants whose implicit female heterosexual stereotypes have been activated no interaction between prime type and target gender was predicted. Because female sexuality is stereotyped as passive, priming for female heterosexual schemas will not relate to the target in the sexually ambiguous vignette; thus this variable is not expected to influence participant's rating of the protagonist and no assimilation effect was expected. No interactions between prime type (female heterosexual or neutral) and target gender was predicted for the other 16 traits on the measure (semantically related and unrelated to female heterosexual). Section II: Explanations for Sexually Ambiguous pehavio; It has been suggested that women who sexually abuse 45 children are viewed as engaging in normal child care behavior or demonstrations of "maternal" affection; whereas men who engage in such behavior are perceived as being sexually inappropriate or abusive. An interaction between prime and target gender was predicted when participants were primed with aggressive, nurturant, and male heterosexual primes. II. III. Hypotheses Regarding Sexually Ambiguous Behavior. When completing sentence stems, participants who were exposed to aggressive primes will explain the male target's ambiguous behavior as more inappropriate (e.g., sexual abuse) than the female target's behavior. No gender difference was expected for participants who have been exposed to neutral primes. When completing sentence stems, participants who were exposed to nurturant primes will explain the female target's ambiguous behavior as more appropriate (e.g., normal child care or maternal affection) than the male target's behavior. No difference was expected for the participants who have received neutral primes. When completing sentence stems, participants who were exposed to male heterosexual primes will explain the male target's ambiguous behavior as more inappropriate (e.g., sexual abuse) than the female target's behavior. No gender difference was expected for participants who have been exposed to neutral primes. 46 IV. No gender difference was predicted for participants who have been exposed to female heterosexual primes or neutral primes on the sentence stem measure. METHOD Dagmar. Students participated in two ostensibly unrelated experiments. In the "first experiment" participants unscrambled short sentences that described behaviors stereotypical of aggression, nurturance, male heterosexuality, or female heterosexuality. Students in the control group unscrambled neutral sentences. The "second study" required participants to read a target paragraph describing a male or female in a sexually ambiguous situation with a child. After a 15-minute filler task, students evaluated the adult target using two dependent measures: a sentence stem measure and a trait rating scale. Subjects Participants were 464 undergraduates, 345 females and 119 males, at Michigan State University participating for course credit. A total of 216 students participated in the aggression and nurturance study (164 females and 52 males). In the male and female heterosexuality study, there were 248 students (181 females and 67 males). Students ranged in age from 17 to 53; the mean age was 19.5. The ethnicity of the participants was as follows: Caucasian 86% (n=399); African American 5.4% (n=25); Native American 1% (n=4); Asian American 3.9% (n=18); Hispanic/Latino 1.7% (n=8); and 47 48 2% (n=10) international students. Materials Demographic Questionnaire Participants completed a short demographic measure (Appendix B) asking their age, gender, and ethnicity. Priming Stimuli The priming stimuli, the first of two independent variables, were designed to activate a participant's unconscious gender stereotypes. The neutral and aggressive priming sentences (see Appendix C) were developed by Banaji et al. (1993). The nurturant and sexual sentences were developed by this author using a procedure similar to the one used by Banaji et al. (1993). See Appendix D for a description of the pilot study used to develop the nurturant and sexual primes. Priming sentences consisted of a three-word sentence which depicted an aggressive, nurturant, sexual, or neutral behavior; a fourth word was added, and the four words were scrambled. An initial, whiCh stood for a name, preceded the four words and was used to prevent gender priming from a specific name. Participants were asked to mark the three words, not including the initial, that formed a sentence. The neutral condition consisted of 45 neutral behaviors; whereas the aggressive, nurturant, male sexual, and female sexual conditions each consisted of 30 target behaviors and 15 neutral sentences. See Appendix C. 49 Target Paragraphs The ambiguous vignette (Appendix E), the second independent variable, was designed to include behavior weakly related to the target traits as well as neutral behaviors (cf. Banaji et al., 1993). One vignette in which the ambiguous behavior could be interpreted as either aggressive, nurturant, or sexual was used. See Appendix D for a description of the pilot work that led to the development of the vignette. The final vignette contained 8 neutral behaviors, and 9 behaviors that could be construed as sexually aggressive or as nurturant (see Story Content Analysis in Appendix E). Typically, half of the behaviors described in a vignette are neutral (cf. Srull & Wyer, 1979) and the other half reflect the primed trait(s) under investigation (of. Erdley & D’Agostino, 1988). The gender of the perpetrator was manipulated by changing the name of the protagonist (Donald or Donna). The child was a girl (Pam) in the Donald vignette and a boy (Pete) in the Donna story so as to avoid the additional stereotypes that a same-gender sexual incident might elicit. W A filler task (Appendix F) was developed that took approximately 15 minutes to complete. Participants were asked to read the first 170 neutral questions (not containing aggressive, nurturant, or sexual content) of the 50 Minnesota Multiphasic Personality Inventory (MMPI) and cross out all the letter "e's." This personality test is no longer in use as it has been replaced by the MMPI-2. Sentence Stems The sentence stems (see Appendix G), the first of two dependent measures, were developed to understand how a participant explained the ambiguous behavior depicted in the vignette. (See Appendix D) for a complete description of the development of this measure.) This measure yielded both a quantitative 3-question scale, and one question provided descriptive data. Sentence Stem Measure. Although initially constructed with six sentence stems, three of these yielded no variability; thus this measure was scored using only three open-ended questions (numbers 3, 5, and 9). All three questions were given a code of 1 for a response that was deemed appropriate, a 0 for an ambivalent response (i.e. those responses that contained both an appropriate and an inappropriate reason), and a -1 for a response that gave an inappropriate reason. Examples of appropriate responses included: "Pam was too young to bathe alone safely," or "Pete may have been having nightmares and was afraid to sleep alone." Examples of inappropriate responses included: "Donald is a psychopathic child molester," or "I felt bad for the child." See Appendix H for a copy of the scoring manual. 51 Scores on the sentence completion measure ranged from +3 to -3. A +3 score indicated that the participant perceived the ambiguous behaviors in the story as appropriate, whereas a -3 score meant that the student interpreted the ambiguous behaviors as inappropriate. This scale had an internal consistency reliability (Cronbach's Alpha) of .60. Descriptive Question. One of the open-ended questions, number 5, also provided descriptive data. Question 5, "Why do you think Donna (Donald) suggested that Pete (Pam) sleep in her (his) bed...", was scored along a continuum: a +3 score indicated that the child's needs were being met; a +2 answer meant both the child's and adult's needs were being met; responses that stated the adult loves or misses the child were scored as +1; students who responded ambivalently received a 0; answers in which the adult's need for company were given as a reason received a -1, references to the adult's mental instability were scored as -2; and participants who's answers indicated that the adult's sexual needs were being met received a -3. See Appendix H for a copy of the scoring manual. Trait Rating Scales The trait rating scales were the second dependent measure. The aggression trait rating scale (Appendix I for copies of the scales) was developed by Banaji et al. (1993). In addition to the target trait, 16 additional traits, half 52 related to the target trait and half unrelated, make up these scales (e.g., the aggressive, nurturant, and sexual trait rating scales). Eight of these 16 words describe positive traits and 8 describe negative traits. The aggression scale used the following 17 traits: aggressive (the target trait); ambitious, confident, independent, and strong-minded (4 related and positive traits); argumentative, belligerent, hot-headed, and stubborn (4 related but negative traits); careful, cultured, educated and neat (4 unrelated and positive traits); and boring, superficial, superstitious, and unhealthy (4 unrelated and negative traits). Similar scales for the nurturant, male sexual, and female sexual traits were constructed by this author using Banaji et al.'s criteria. The 8 positive traits were chosen and equated based on ratings of likability (Anderson, 1968) and word frequency (Kucera & Francis, 1967), as were the 8 negative traits. The nurturant scale used the following 17 traits: nurturant (the target trait); dependable, unselfish, friendly, and kindly (4 related and positive traits); worrier, oversensitive, bossy, and nosey (4 related but negative traits); open-minded, interesting, humorous, and cultured (4 unrelated and positive traits); and unoriginal, superstitious, uncivil and cowardly (4 unrelated and negative traits). 53 The male sexuality scale used the following 17 traits: sexual (the target trait); vigorous, confident, untiring, and tender (4 related and positive traits); crude, vulgar, dominating, and possessive (4 related but negative traits); productive, rational, easygoing, and witty (4 unrelated and positive traits); and superficial, boring, envious and superstitious (4 unrelated and negative traits). Finally, the female sexuality scale used the following 17 traits: sexual (the target trait); vivacious, eager, active, and warm (4 related and positive traits); fickle, vain, inhibited and passive (4 related but negative traits); cultured, educated, interesting and curious (4 unrelated and positive traits); and superstitious, petty, impractical, and forgetful (4 unrelated and negative traits). Participants rated the protagonist on a 10-point Likert scale from 1 (not at all) to 10 (extremely). A high score indicated that the student thought the trait was representative of the target individual; a low score meant the participant did not find the trait representative of the protagonist. The target trait (aggressive, nurturant or sexual) was analyzed alone; the other 16 traits on each scale were analyzed in groups of four. For example, on the aggressive trait rating scale the mean of the four related and positive traits (i.e. ambitious, confident, independent and strong-minded) was used to analyze the participants response to that group of four words. 54 Procedure The priming and vignette experiments were administered in small groups. Each participant was assigned to one of two studies. Each study contained two experiments; each experiment was a 2 X 2 between-subjects design: Experiment 1A: 2(Prime Type: aggressive or neutral) X 2(Target Gender: male or female protagonist); Experiment 18: 2(Prime Type: nurturant or neutral) X 2(Target Gender: male or female protagonist); Experiment 2A: 2(Prime Type: male heterosexual or neutral) X 2(Target Gender: male or female protagonist); Experiment 28: 2(Prime Type: female heterosexual or neutral) X 2(Target Gender: male or female protagonist). The first 216 participants were assigned to Study 1 and then randomly assigned to one of six prime type X target gender conditions (the same control group was used for Experiments 1A and 13). The second 248 participants were assigned to Study 2 and then randomly assigned to one of the six prime type X target gender conditions (Experiments 2A and 28 shared the same control group). Upon entering the room, students were told that they would be participating in two separate "experiments" because both were rather short. In addition, the "experiments" were administered by two individuals, the materials in the two "experiments" were printed using different fonts, and two very different consent forms (Appendix J) were used to aid the deception. This deception was necessary for the priming 55 manipulation to succeed: participants could not know that they were being primed or the priming would no longer be unconscious. The first experimenter (an undergraduate research assistant) introduced herself and explained that she was working for a professor who was designing a new neuropsychological test for the assessment of dementia. In order to standardize this test, the professor needed to collect data from a variety of age groups. The research assistant explained the consent form, and showed students the "neuropsychological test." She explained that it was. important to time how long it took students to complete the test but that accuracy was important as well because the researchers were interested in how long it took normal, non- demented people to complete the test, and how many mistakes were commonly made. After handing out stop watches, the priming materials were distributed and students began working. In the first study, participants received one of three prime types (or "neuropsychological tests"): aggressive, nurturant, or neutral. In the second study, students received either male heterosexual, female heterosexual, or neutral primes. After the entire group was finished unscrambling the primes, the research assistant collected the materials, and the second experimenter began her "experiment." In the second "experiment," administered by the author, 56 students were told that they would be participating in an educational experiment--a reading, writing, and concentration task. A packet of materials containing the demographic questionnaire, vignette, filler task, and dependent measures was shown to students, and the various activities were briefly described. It was explained to participants that because this experiment took individuals different amounts of time to complete, after they had finished they could leave one at a time to be "checked out." (This was necessary, as each subject had to be individually debriefed to ascertain if they saw a connection between the two "experiments," thus a group debriefing was not possible.) Students completed a second consent form, and then began working on the packet. Participants received a vignette with either a male target or a female target engaging in sexually ambiguous behavior with a six year old child. After reading the vignette, students spent 10 to 15 minutes completing a concentration task (the filler task). The sentence completion measure was written to correspond to the male or female protagonist. The trait rating scale corresponded to the type of primes the students had originally received; thus, those receiving aggressive primes received an aggressive trait rating scale, etc. Students in the control group received both trait rating scales for their study (the presentation was counterbalanced) and 57 served as the control group for both prime groups. Thus, in the first study the controls received an aggressive and a nurturant trait rating scale; in the second study they received a male and a female heterosexual trait rating scale. After completion of the "second experiment," participants were individually debriefed to ascertain if they realized there was a connection between the "first" and the "second experiments." After this debriefing, any questions students may have had were answered, the nature of the experiment was explained to them, and they were given a feedback sheet (Appendix K). In the first study, no participants made a connection between the priming task and the education test. But in the second study, 30 students made an accurate connection between the two "experiments"; 25 participants receiving the male heterosexual primes saw a sexual connection in both of the "experiments," and 5 students who had the female heterosexual primes stated that both "experiments" had a sexual content. Because the priming was no longer unconscious for them, these 30 individuals were excluded from the final data analyses. The data was coded by the author and two undergraduate research assistants who were blind to the hypotheses. An interrater reliability of .98 was calculated for the three coders using a subsample of 50 protocols (10.7% of the data). RESULTS mm This dissertation consisted of two studies; within each study there were two experiments that shared a control group. The first study examined the effects of aggressive, nurturant, or neutral primes on subsequent ratings of a target individual; the second study explored the effects of male heterosexual, female heterosexual, or neutral primes on judgements of the protagonist. Thus, there were a total of four 2(Prime Type:active or neutral) X 2(Target Gender: male or female protagonist) experiments. Prime type and gender of the protagonist were the two independent variables. The two dependent measures were a trait rating scale and a sentence stem measure. Only results at the .05 level or greater were considered significant. The results are presented as four experiments, not as two studies, because each prime type is analyzed independently, and there are four prime types. The two studies refer to the shared control group, not to a comparison between aggressive and nurturant priming, or male and female heterosexual priming. For each of the four prime types (i.e. aggressive, nurturant, male heterosexual and female heterosexual), the analyses of the trait rating scale 58 59 will be discussed; then the results of the sentence stem measure will be presented. As discussed previously, each trait rating scale consisted of 17 traits that were judged by the participants as not at all representative or extremely representative of the target individual using a 10-point Likert scale. For each of the trait rating scales, five two-way analyses of variance (ANOVAs) (Prime Type X Target Gender) were conducted. The first ANOVA was an analysis of the target word (i.e., aggressive, nurturant, male heterosexual or female heterosexual); the second was an analysis of the four positive words related to the target word; the third analysis was of the four negative words related to the target trait; the fourth ANOVA analyzed the four positive words unrelated to the trait of interest; and, finally, the fifth analysis was of the four negative words unrelated to the target trait. The sentence stem measure, as described previously, computed participants' responses to three open-ended questions. Scores ranged from +3 to -3. Participants who gave vague or unscorable answers to the open-ended questions were excluded from the analyses, thus the number of participants is lower for this dependent variable than it is for the trait rating scales. The sentence stem measure was analyzed using a two-way ANOVA (Prime Type X Target Gender) for each prime type. Thus, there are four analyses of the 60 sentence stem measure, one for each of the prime types (i.e., aggressive, nurturant, male heterosexual, and female heterosexual). After the analyses of the trait rating scale and sentence stem measure have been presented, a brief summary of the qualitative data from Question 5 on the sentence stem measure will be given. Results for the Aggressive Prime Condition An interaction between prime type and target gender was predicted for the aggressive prime condition; male targets were hypothesized to be evaluated as more aggressive than) female targets. For the target word aggressive, analysis of variance revealed a marginally significant Prime Type X Target Gender interaction, 2(1, 140) = 3.76, p < .054; after exposure to aggressive primes, participants' ratings of the male targets increased and ratings of the female targets decreased. Because the interaction was almost significant, post-hoc comparisons of the means were performed. One-way ANOVAs indicated that there was no difference between the male and female target means in the neutral prime condition (M = 4.19 vs. M = 4.34; F(1, 70) = .05), but there was a significant difference in the aggressive prime condition. Male targets had higher means than female targets (M = 5.03 vs. M = 3.34; £(1, 70) = 6.39, p < .01). Comparisons between the two male target means and the two female target means were not 61 significant (M = 5.03 vs. M 4.19; £(1, 72) = 1.54, and M = 4.34 vs. M = 3.34; F(1, 68) 2.26). See Table 2 and Figure 1 in Appendix A. In addition, contrary to the hypothesis, four main effects for gender were observed on the traits semantically related and unrelated to aggressive. For the two groups of traits that were positive, the female target was rated higher than the male target (M = 6.08 vs. M = 5.42, and M = 6.41 vs. M = 5.67). For the two groups of traits that were negative, the male protagonist was rated higher than the female protagonist (M = 3.65 vs. M = 2.94, and M = 4.41 vs. M = 3.56). See Table 2 in Appendix A. When completing the sentence stem measure, it was hypothesized that participants who had been exposed to aggressive primes would explain the male target's ambiguous behavior as more inappropriate than the female target's behavior. The predicted interaction for the aggressive prime condition was not significant. A significant main effect for gender was observed, F(1, 100) = 15.69; p < .001. The female target's behavior was rated as more appropriate than the male target's behavior (M = 1.56 vs. M = .14). See Table 3 in Appendix A. Results for the Nurturant Prime Condition An interaction between prime type and target gender was hypothesized for the nurturant prime condition; the female target was expected to be judged as more nurturant than the 62 male target. The predicted interaction for the target trait nurturant was not significant. No significant interactions for the other traits semantically related or unrelated to nurturant were observed. A significant main effect for prime type was observed for positive traits semantically related to nurturant, 3(1, 140) = 15.39, p < .001. Participants who received the nurturant primes rated the targets higher than the individuals who received control primes on the positive traits semantically related to nurturant (M = 8.10 vs. M = 7.03). See Table 4 in Appendix A. For the sentence stem measure, it was hypothesized that students who had been exposed to nurturant primes would explain the female target's ambiguous behavior as more appropriate than the male target's behavior. An ANOVA revealed a significant Prime Type X Target Gender interaction, F(1, 96) = 4.33; p < .05. The interaction was in the opposite direction from the prediction; after exposure to nurturant primes participants', ratings of the appropriateness of the male target's behavior increased and ratings of the appropriateness of the female target's actions decreased. Post-hoc comparisons of the means were performed. One-way ANOVAs indicated no difference between the male and female targets' means except in the neutral prime condition (M = .19 vs. M = 1.42, 3(1, 51) = 5.93); p < 63 .01. See Figure 2 and Table 3 in Appendix A. Results for the Male Heterosexual Prime Condition An interaction between prime type and target gender was predicted for the male heterosexual prime condition; it was hypothesized that participants would evaluate the male target as more sexual than the female target. The predicted interaction for the target trait male heterosexual was not significant. No significant interactions for the other traits semantically related or unrelated to male heterosexual were observed. A significant main effect for target gender was observed for positive traits semantically related to male heterosexuality 3(1, 141) = 4.74, p < .05. Participants who read a vignette involving a female target rated the female target higher than individuals who received a story with a male target (M = 6.12 vs. M = 5.61) on the positive traits semantically related to male heterosexuality. See Table 5 in Appendix A. When completing the sentence stem measure, it was hypothesized that participants who had been exposed to male heterosexual primes would explain the male target's ambiguous behavior as more inappropriate than the female target's behavior. The predicted interaction for the male heterosexual prime condition was not significant. In addition, no significant main effects were observed. See Table 3 in Appendix A. 64 Results for the Female Heterosexual Prime Condition Participants who had been primed with female sexual primes were not expected to judge female targets as more sexual than male targets. As hypothesized, the interaction for the target trait female heterosexual was not significant. No significant interactions for the other traits semantically related or unrelated to female heterosexual were observed. In addition, no significant main effects were observed for the target trait or the other traits semantically related or unrelated to female heterosexuality. See Table 6 in Appendix A. For the sentence stem measure, no target gender differences were expected for the students who had received female sexual primes or neutral primes. As predicted, the interaction for the female heterosexual prime condition was not significant. In addition, no significant main effects were observed. See Table 3 in Appendix A. Qualitative Results As previously described, Question 5 on the sentence stem measure was also scored using a categorical system. See Table 7 for a summary of Study 1 and 2 responses to the 7 categories for Question 5. What is most striking about these results are the number of responses to Category -1 (adult's need for company) and -3 (adult's sexual needs). For Study 1, more male targets received a -3 score than female targets. For 65 Study 2, many more female targets received a -1 response than male targets, but more male targets received a -3 score than did female targets. DISCUSSION This experiment is the first attempt to provide a theoretical explanation for why sexually abusive men and women are judged differently by observers. Two prior studies (Broussard et al., 1991; Finkelhor & Redfield, 1984) have demonstrated that women are judged as less harmful when sexually abusing a child, and as less representative of child sexual abusers, as compared to men. Surveys of mental health professionals also indicate that female perpetrators are perceived as less harmful than male perpetrators (Eisenberg, Owens, & Dewey, 1987; Wagner, Aucoin, & Johnson, 1993). Other researchers have suggested that women's sexually abusive behavior is misperceived as nurturant, and the sexual nature of the acts are masked (e.g., Banning, 1989: Goodwin & DiVasto, 1979; Groth, 1979a; Plummer, 1981); although, no empirical data has been presented to support these suggestions. The present investigation sought to explain these observed differences using a theory of implicit gender stereotyping. Four gender stereotypes were hypothesized to be related to perceptions of ambiguous child sexual abuse: male aggression, female nurturance, and male and female heterosexuality. The findings will be discussed as they relate to implicit gender stereotyping and to perceptions of 66 67 child molesters. First, the interaction on the aggressive trait rating scale will be discussed, then the main effects for gender across experiments will be addressed, and, finally, the results for the nurturant prime experiment will be presented. Aggressive Trait Rating Scale Interaction In the aggressive prime experiment, participants in the experimental condition received primes intended to activate their unconscious stereotypes of male aggression. It was hypothesized that participants who received aggressive primes and the male target would rate the target as more aggressive compared to those who received aggressive primes and the female target. The interaction was marginally significant. Post-hoc comparisons revealed that mean ratings of the male target were higher than those of the female target in the aggressive prime condition, but there were no differences between the other means. The primed trait (i.e., aggressive) was applicable only to the male target, not to the female target. This finding lends some support to the results of other studies that have documented the necessity of social category applicability for unconscious priming to have an effect (e.g., Banaji et al., 1993). There is no support for assimilation or contrast effects. Assimilation effects are expected when the primed trait is applicable to the social category of the target and 68 this leads to a significant increase in the experimental group mean compared to the control group mean (Banaji et al., 1993). Contrast effects are more unusual, indicating that the primes are negatively associated with the target's social category (Banaji et al., 1993; Srull & Wyer, 1980). Contrast effects produce a significant decrease in the experimental group mean compared to the control group mean. In this experiment, priming had no effect on the ratings of the female target because women are not stereotypically associated with aggression. If a contrast effect had been found, it might have indicated that women are negatively associated with aggression in a sexually ambiguous situation. Using the exact same primes but a very innocuous, non- sexual vignette, Banaji et al. (1993) found an assimilation effect for male targets (ratings increased after exposure to aggressive primes). The interaction between prime type and target gender was not significant, and there was no significant difference between ratings of the male and female target in the aggressive prime condition. 'These authors speculated that, in their experiment, the target may have been perceived as assertive and not aggressive due to the weakness of the stimulus materials. This may have resulted in their nonsignificant interaction. The marginally significant interaction in this experiment could also have been produced by weak primes that only stimulated 69 stereotypes of assertiveness. Alternatively, Banaji et al. suggest that a counterstereotype of an aggressive female may have produced the smaller effect. Recall that they found no significant difference between ratings of male and female targets in the aggressive prime condition. The results from the present study indicate that there is probably no counterstereotype of an aggressive female when the situation involves ambiguous sexual abuse because male targets were rated as significantly more aggressive than female targets in the aggressive prime condition. The finding of a contrast ' effect would have provided strong support for the lack of an aggressive female counterstereotype. If replicated and found to be significant, the above marginally significant interaction would demonstrate that child sexual abuse is perceived as an aggressive behavior, but only for male targets. Banaji et al. (1993) have noted that stereotyping is more responsive to recent experiences than previously thought. In situations in which individuals are discussing child sexual abuse, or listening to reports of sexual abuse in the media, hearing the phrase "child sexual abuse" may serve as a prime to stimulate stereotypes of men aggressively preying on children and molesting them. If the abusive behaviors are somewhat questionable or ambiguous, male perpetrators might be much more likely to be perceived as aggressive and abusive because the activated 7O stereotype would apply only to males. This could result in the underreporting of female sexual abusers when an individual's aggressive stereotypes have been activated. The hypothesized interactions for the other three primed stereotypes (i.e., nurturant, male heterosexual, and female heterosexual) were not significant. The findings for the nurturant prime experiment will be discussed in the next section. Two explanations for the lack of a significant interaction on the male and female heterosexuality trait rating scales are possible: 1) the priming manipulation could have been ineffective, in which case the hypotheses were not tested, or 2) priming with male or female heterosexual stereotypes was not applicable to the targets in the sexually ambiguous vignette. First, priming experiments generally activate unidimensional traits such as honest or mean (Erdley & D'Agostino, 1988), aggressive or dependent (Banaji et al., 1993), hostile or kind (Srull & Wyer, 1979), and kind or shy (Bargh et al., 1986). The above examples are specific traits that can be described with one word synonyms. In this study, male heterosexuality was stereotyped as dominance, and female heterosexuality as passivity, but, it is likely that heterosexuality is much more complicated and multidimensional. Unlike traits such as kind or aggressive, male and female heterosexuality are multifaceted concepts. They may not readily lend themselves to priming, both 71 because they are too complicated and may even contain inherent contradictions. Honesty has one basic meaning; however, male heterosexuality can mean many different things, including dominance, tenderness, and gentleness. Second, research on unconscious priming has demonstrated that priming has no effect if it is not relevant to the target (e.g., Banaji et al., 1993; Higgins, Rholes, & Jones 1977). No interaction was expected for the female heterosexual primes because female sexuality is stereotypically perceived as passive (Hite, 1993). Ambiguous sexually abusive behavior was not expected to be perceived as passive, thus priming for sexually passive stereotypes would not be relevant to male or female targets engaging in ambiguous child sexual abuse. Given that male heterosexuality is stereotyped as containing elements of dominance and power (Hite, 1993), priming with male heterosexual stereotypes should be applicable to male targets in a sexually ambiguous situation involving a child. From this study, one might conclude that the adult target's ambiguous sexual contact, as depicted in the vignette, is not perceived as male heterosexual behavior, because priming for male heterosexual stereotypes had no effect. This suggests that the primes are not relevant to the targets or are not relevant to the vignette. Child sexual abuse may be understood by participants as a sexual perversion and a crime of aggression, thus, 72 activating stereotypes of male heterosexuality would not be applicable to the target in the vignette. Based on this study, unconscious stereotypes of male aggression offer the best explanation for perceived differences between male and female child molesters. Main Effects for Gender Four unpredicted main effects for gender were observed on the aggressive trait rating scale; main effects were observed on traits semantically related and unrelated to aggressive. Specifically, the female target was rated higher when the traits were positive, and the male target was rated higher when the traits were negative. A significant main effect for gender was observed on the male heterosexual trait rating scale. The female target was rated higher on positive traits semantically related to male heterosexuality, and a trend toward significance was observed for higher ratings of the female target on positive traits unrelated to male heterosexuality. Finally, on the sentence stem measure in the aggressive prime experiment, the female target's behaviors were rated as significantly more appropriate than the male target's actions. In experiments without priming, researchers have found that female child sexual abusers are rated as less abusive than are males when in the exact same situation (Broussard et al., 1991; Eisenberg et al., 1987; Finkelhor & Redfield, 1984; Wagner et al., 1993). Although, the main effects for 73 gender found in the present study substantiate these observations, the question remains: Why are women perceived more positively than men when in a sexually ambiguous situation with a child? An explanation for these main effects is that they are an expression of conscious (as opposed to unconscious) gender stereotyping. 0n the traits that are evaluatively negative, students' might rate the male target higher as a means of expressing displeasure toward the male in the vignette. They may be rating the female lower on these traits because the woman target is perceived in a more- positive light in the vignette. The reverse would be true for the positive traits; the woman target is rated higher because she is perceived as more appropriate, and the male target is rated lower because he is evaluated as less positive. This explanation is supported by the results on the sentence completion measure. The female target's ambiguous behaviors were rated as more appropriate than the male target's ambiguous actions. A comparison of the means reveals that the female target's behaviors were rated as appropriate, whereas the male target's actions were rated as ambivalent. It has been suggested, but no empirical data has been presented, that women's sexual perversions are denied or ignored (Welldon, 1988; 1991). In this study's vignette, participants may have ignored the possible sexual nature of 74 the situation in the case of women protagonists, and this may have led to more positive ratings for female targets. An examination of the qualitative results from Question 5, on the sentence stem measure, lends empirical support to this statement. When asked to explain why a target suggested the child sleep in his or her bed, participants responded that the female target wanted the child for company or companionship, but the male target invited the child into bed for sexual reasons. Although both responses were rated as inappropriate, the female target's behavior was perceived as more innocuous than the male target's.' The possible sexual nature of the activity was generally ignored for the female targets. Nurturant Prime Experiment The results from the nurturant prime experiment are presented separately as they seem to be an aberration; the use of positive priming stimuli may have resulted in the stimulation of positive caretaking stereotypes. In the nurturant prime experiment, participants in the experimental condition received primes intended to activate their unconscious stereotypes of female nurturance. It was hypothesized that there would be an interaction with participants who received nurturant primes and female targets rating the targets as more nurturant compared to those receiving nurturant primes and male targets. This hypothesis was not supported. An unexpected significant 75 main effect for prime type on traits semantically related to nurturant (i.e., dependable, unselfish, friendly, and kindly) indicated that priming did have an effect, but the effect was not gender specific, nor was the effect specific to the trait nurturant because the main effect was observed on positive traits related to nurturant. On the sentence stem measure, it was hypothesized that nurturant priming would lead to an interaction such that students who received the nurturant primes and a female target would rate the female target's behavior as more appropriate. This hypothesis was not supported. An interaction in the opposite direction from that which was hypothesized was observed; nurturant priming led to an increase in appropriateness ratings of the male target and a decrease in appropriateness ratings of the female target. A pppp-ppg comparison of means revealed that the only significant difference in ratings was between the male and female targets in the neutral prime condition. Although the interaction was significant, there was no support for assimilation or contrast effects. Nurturant priming had no effect on ratings of female targets on the sentence stem measure. This interaction and main effect were unanticipated; nurturant priming was not expected to be relevant to male targets as there is no social category applicability. In this study, the primes appear to have activated generic, 76 unconscious caretaking stereotypes. Because the nurturant primes worked non-specifically; that is, they were not dependent on a target's social category, participants may have had stereotypes of both caretaking males and females. The students in this study, with a mean age of 19.5, were raised at a time when fathers were encouraged by society to take a more active role in caring for their children. This sample could possess stereotypes of both male and female caretakers. The activation of caretaking stereotypes could account for the main effect, which was not gender specific but dependent on the type of prime received. On the sentence stem measure, it is possible that with "nurturant priming" the male targets' ambiguous behavior was interpreted through a caretaking prism. Perhaps the "nurturant primes" served to "normalize" the male target's ambiguous actions in contrast to the judgements of the students who received neutral primes. In the vignette, the male target performed behaviors (e.g., picking the child up from school, taking the child to the park, and feeding the child dinner) that were unequivocally caretaking. With the activation of caretaking stereotypes, the male target's sexually ambiguous behaviors (i.e., kissing the child, helping the child bathe, and putting the child to sleep in his bed) could be perceived as appropriate caretaking activities. An alternative explanation for the above findings is 77 that the positive nature of the nurturant priming stimuli activated a global, positive affective schema, and this was transferred onto the ratings of the targets (see Bargh et al., 1986; Erdley & D'Agostino, 1988). Because nurturance is an unequivocally positive trait it was not possible to generate examples of nurturant behavior that were negative when creating the priming stimuli. The development of the other three priming stimuli (i.e., the aggressive and male and female heterosexual primes) involved choosing negative (as rated by pilot subjects) examples of the stereotypes in question to insure that the most extreme examples of the behavior were being used (M. R. Banaji, personal communication May 9, 1995). When creating the nurturant priming stimuli, very few examples of nurturant behaviors were judged by pilot participants as negative. Thus, it was deemed necessary to choose positive examples of nurturant behavior for the primes. (See the Pilot Study in Appendix D for a description of the formation of the priming stimuli.) The literature on the transfer of priming to other semantically related or evaluatively related traits is inconclusive (Erdley & D'Agostino, 1988). Banaji et al. (1993) designed their trait rating scale to test for the possibility of priming extending to other traits. Only the primed stereotype demonstrated a prime type by gender interaction. Devine (1989) found that priming effects transfer only to semantically related trait dimensions--not 78 to evaluatively related traits. Erdley and D'Agostino (1988) also found no support for the transfer of priming effects due to the stimulation of affective responses. Activating negative traits did not result in higher ratings on other negative traits unrelated to the primed trait. Yet, Srull and Wyer (1980) found that priming effects only generalized to traits that are evaluatively related but descriptively (semantically) unrelated to the primed trait. In the nurturant experiment, the priming affected traits that were both semantically related and evaluatively positive. Higgins, Rholes, and Jones (1977) found that using semantically related and evaluatively positive traits does result in higher ratings of desirability of the target than does the use of semantically related but negative traits. This is only true when the primed traits are applicable to the target individual (Higgins et al., 1977). Thus, it appears likely that the "nurturant primes" were relevant to both the male and female targets because of the existence of male and female caretaking stereotypes. In addition, the positive nature of the priming stimuli may have also contributed to male and female targets being rated higher on the positive traits related to nurturance, compared to the ratings of targets in the neutral prime condition. The sentence stem measure could also be assessing the effects of both the "nurturant priming" having stimulated stereotypes of caretakers and a positive 79 cognitive schema. This would explain the interaction; the male target was rated as more appropriate after exposure to the nurturant primes. The primes stimulated caretaking stereotypes, which applied to male targets, and the primes stimulated a global affective reaction which resulted in more positive ratings for male targets. The priming stimuli seem not to have captured the essence of female nurturant behavior adequately, because the effect was neither gender nor trait specific. Thus, the nurturant hypothesis was not actually tested. Stronger or more accurate priming stimuli would have to be developed to test this hypothesis. A review of the nurturant primes (see Appendix C) does not indicate that they were strikingly inadequate, but perhaps they were too universal (i.e., not gender specific enough) and this served to stimulate more general caretaking stereotypes. In other words, it may not be possible to activate stereotypes of female nurturant behavior with the priming method that was used in this study. The short sentences necessary for this methodology may be too general to prime stereotypes of nurturant females; generic caretaking stereotypes may always be activated. The use of a subliminal priming method, in which the priming stimuli are presented outside of the participant's conscious awareness, would offer the opportunity to use very specific words for priming (Erdley & D'Agostino, 1988). The use of words such 80 as maternal, birth, nurse, motherly, succor, lactation, and pregnant, which relate only to females, might be much more likely to activate specific stereotypes of nurturant women. The use of such specific words in the priming method used in this study would probably result in the primed trait or stereotype becoming conscious. There is some evidence that the use of a priming method in which participants consciously attend to the priming task (the method used in this study) produces a greater priming effect for negative than for positive trait categories (Srull & Wyer, 1979). Thus, in future research, it may be more effective to use a subliminal priming method to activate positive trait categories, such as female nurturant behavior. Because it appears that female nurturant primes were not activated, no information is provided to support or refute the suggestions made in the literature that women's sexually abusive behavior is erroneously perceived as nurturant behavior, and the sexual nature of the actions are masked (e.g., Banning, 1989; Goodwin & DiVasto, 1979: Justice & Justice, 1979; Kempe & Kempe, 1984). The differential perceptions of men and women's sexually abusive behavior may be a result of both the stimulation of unconscious aggressive stereotypes, as previously discussed, and unconscious expectations that women are behaving appropriately with children. Further testing of the effects of unconscious nurturant stereotypes, using a subliminal 81 priming method, is suggested. Limitations of the Study There are a number of possible methodological limitations with this study. First, there were methodological problems with the sentence stem measure. Second, the control groups for Study 1 and Study 2 were significantly different on this dependent measure. Third, the behaviors used in the vignette were not rated for their applicability to the primed stereotypes. Fourth, the participants were predominantly Caucasian females. The sentence stem measure did not yield as much data as was expected. It was anticipated that students' thoughts and opinions about the ambiguous behaviors would be expressed on this measure but, generally, the questions were answered exactly as they were presented in the vignette. Although presented as a generic educational experiment, students may have thought of it as a "reading comprehension test" in which one strives to answer the questions as accurately as possible. If the vignette had been taken away from students after they had finished reading it, they may have been more forthcoming with their own perceptions. Participants were routinely observed turning back to the vignette to answer the questions. If the experiment had been described as one which required students to form an impression of the target, rather than as an educational experiment, they may also have been less inhibited with 82 their responses. Because of these limitations, this measure did not provide detailed information about the participants' understanding of the ambiguous behaviors. For example, although the data allowed the experimenter to rate the appropriateness/inappropriateness of the ambiguous behaviors, ratings regarding whether or not female target's sexual actions were misperceived as maternal affection or as nurturant behavior could not be made. In addition, because participants often just reiterated the vignette when answering the questions, answers to only three questions had any variability in their responses. Thus, this scale consisted of only three items and had a relatively low internal consistency rating which may have created error variance and resulted in the scale not providing a true measure of participants' ratings of the appropriateness/inappropriateness of the target's behavior. Taking the methodological limitations of the sentence stem measure into consideration, a final point concerning the control groups for this measure will be made. An examination of the control groups, from the first and the second studies, indicated that they were not equivalent. Recall that two control groups (with either a male or female target) were used in the aggressive and nurturant study, and two control groups (with either a male or female target) were shared in the male and female heterosexuality study. In the aggressive and nurturant study, the means for the two 83 neutral prime groups were not equivalent, 3(1, 51) = 5.93, p < .01; the female target’s behavior was rated as more appropriate than the male target's. In the male and female heterosexuality study, the means for the two neutral prime groups were not significantly different, 3(1, 50) = 2.21. Because the vignette and the neutral priming stimuli were identical in both studies, one should not expect to see a significant difference between the two sets of control groups in the first and second study. Studies have consistently documented perceived differences between male and female child molesters using a variety of research and sampling techniques (Broussard et al., 1991; Eisenberg et al., 1987; Finkelhor & Redfield, 1984; Wagner et al., 1993). These studies suggest that there is a difference between ratings of appropriateness of male and female targets. As previously discussed, this gender difference may reflect a denial of the sexual aspects of the female target's ambiguous behaviors, and result in higher ratings of appropriateness. No rating differences between target gender on the sentence stem measure in the male and female heterosexual experiments may have been observed because the control group in Study 2 did not rate male and female targets differently. This observed difference raises an interesting concern: If the control group, without any priming, already rates the male and female targets differently, then the effects of 84 priming alone cannot explain these differential ratings. Yet, on the aggressive trait rating scale no difference was observed for the control groups' ratings of male and female targets for aggression. The difference for the control groups on the sentence stem measure may well reflect the methodological limitations of this measure. True control group differences could also suggest that the differential perceptions of male and female child molesters is multifaceted. As previously discussed, both unconscious aggressive stereotypes and conscious gender stereotypes might influence perceptions of male and female targets. In addition, participants chronically accessible traits will influence their ratings of a target without the necessity of priming. This will be discussed more fully in the final section. The vignette construction technique may have also contributed to the weak findings in this study. Vignettes used in priming experiments are designed, in part, by having participants rate the behaviors that will be used in the vignette on a scale indicating how representative the behaviors are of the traits that will be primed (Banaji et al., 1993; Srull & Wyer, 1979; Srull & Wyer, 1980). For example, in a dependent priming experiment, behaviors indicative of dependence are rated by pilot participants for how dependent they are, and, then, only behaviors weakly related to dependence are chosen for the vignette. 85 The vignette used in this project was not constructed in this manner because the behaviors in the vignette were weakly related to a number of traits (i.e, aggression, nurturance, sexuality). It was believed that it would be too confusing to ask participants to rate behaviors for their representativeness on a number of traits. Instead, the author chose a number of behaviors that were implicated in the literature as actions performed by women (such as bathing a child) in which the sexual nature of the actions could be misperceived as caretaking or nurturant acts. The number of behaviors in the vignette were balanced for their nurturant, aggressive/sexual, or neutral content, but pilot participants were not asked to rate the behaviors for their representativeness. Thus, the behaviors in the vignette may have been strongly representative of neutral and/or nurturant behaviors and inadequately depicted sexual behaviors. Ideally, the behaviors in the vignette should weakly reflect the primed traits. A final possible limitation of this study is the overwhelming homogeneity of the participants, who were predominantly females. The sample of males was too small to make any comparisons possible between male and female participants' ratings. Studies have documented that women tend to rate child abuse, including sexual abuse, as more serious than do men (Broussard et al., 1991; Dukes & Kean, 1989; Eisenberg et al., 1987). The women participants might 86 have rated the behaviors in the vignette as more inappropriate then the male students. The second demographic limitation is the ethnicity of the participants; the students in this study were overwhelmingly Caucasian. The sample of minority participants was too small to make any comparisons possible between Caucasian and minority perceptions of adults in a sexually ambiguous situation with a child. Harrison-Speake and Willis (1995) found racial differences between respondents when judging appropriateness of touch among family members. The results from this study should not be generalized to the population at large. Eggppp Directions for Research The results from this study are best understood as preliminary; replication and extension of this work is necessary. Although little support was found for implicit gender stereotypes differentially influencing perceptions of male and female child molesters, the marginally significant interaction in the aggressive prime experiment warrants further study. A replication of this study with changes and additions is suggested. First, the primes should be tested in an unrelated vignette with social category applicability to insure that they are indeed activating the stereotypes of interest. The use of a subliminal priming technique, in which very specific words could be used to activate the 87 traits, might prove to be more successful, especially for the nurturant prime condition. A larger sample, more representative of the general population, should be used. The vignette should be designed using a standardized method to insure that no one trait is over represented. As described previously, this could be accomplished by asking students to rate a list of behaviors for their aggressive, nurturant, sexual, or neutral content, and then choosing the behaviors that are rated as weakly related to the traits of intereét. The sentence stem measure could be better designed to assess more accurately participants' perceptions of the ambiguous behavior by rewording the stems and giving students instructions to form an impression of the target individual. Finally, it would be interesting to measure the effects of chronic trait accessibility as they apply to evaluations of male and female sexual abusers. Research has demonstrated the additive nature of priming in individuals who have chronically accessible constructs (Bargh et al., 1986). Chronic accessibility is present across situations and is not dependent on priming to influence how individuals perceive and interpret their social reality (Bargh et al., 1986). Thus, if individuals have chronically accessible cognitive constructs of a child sexual abuser or nurturing women, priming would not be necessary for these constructs to shape their perceptions of a sexually ambiguous 88 situation. Individuals' chronically accessible constructs might explain why the control groups in Study 1 and Study 2 differed. If chronic accessibility was measured prior to the beginning of the study, students high in a trait of interest could be randomly assigned to groups, and the effects of both chronic accessibility and priming could be measured. The factors producing differential perceptions of male and female child molesters might be the result of the interaction of a number of stereotypes and the chronic accessibility of those stereotypes in an individual. For example, one could examine the effects of power on individuals' perceptions of child sexual abusers. If child sexual abuse is best perceived as a crime of aggression, not a sexual act, individuals' stereotypes of power may also contribute to the formation of gender specific judgments. Because men are generally perceived as more powerful or agenic than women (Bergen & Williams, 1991; Gerber, 1991), men might also be more likely to be perceived as child sexual abusers when an individual's implicit stereotypes of power have been activated. Conclusion The near significant interaction in the aggressive experiment (in the direction predicted) suggests that the activation of aggressive stereotypes might best explain the observation that women are less likely than men to be 89 perceived as child molesters. The priming of unconscious stereotypes of male heterosexuality had no effect on participants' subsequent judgements in a sexually ambiguous situation. This suggests that stereotypes of male heterosexuality are not relevant to targets in a sexually ambiguous situation with a child. In this study, female targets were generally rated more positively than male targets, supporting observations from other studies on child sexual abuse. Why women are judged less harshly in a sexually ambiguous situation with a child is still unclear, but, it has been suggested, and there is some support from data from the present study, that the sexual aspects of the female target's behavior are ignored. No support was found that a woman's sexually abusive behavior is misperceived as nurturant, but the nurturant hypothesis was not adequately tested. The results for the nurturant experiment are unexpected and may have resulted from the activation of caretaking stereotypes and positive affective schemas. This work was the first attempt to provide a theoretical explanation for the empirical observation that sexually abusive men and women are evaluated differently. Because of the limitations of this study, the effects of unconscious priming of gender stereotypes have not been adequately tested as they apply to perceptions of male and female child sexual abusers. The near significant 90 interaction for the aggressive prime experiment warrants replication with methodological changes to further examine if unconscious gender stereotypes offer a theory to understand why men and women are perceived differently when sexually abusing a child. APPENDICES APPENDIX A Tables and Figures 91 Table 1 Percentage of Women Child Sexual Abuse Perpetrators: Incarcerated Peppetrators Study Date Sample % Women Abusers Musk & Gallagher 1985 USA 1.65% O'Connor 1987 England .2% Rowan, Rowan & Langelier 1990 Vermont and New Hampshire 1.5% Male Victims of Sexual Abuse in Prison Study Date Sample % Women Abusers Allen 1991 convicted child molesters 45% Burgess et al. 1987 male rapists 40% Condy et al. 1987 convicted rapists 57% convicted child molesters 37% other convicts 47% Groth 1979b sexual assault convicts 41% Petrovich & Templer 1984 convicted rapists 59% Agencv Cases Study Date Agency % Women Abusers DeFrancis 1969 American Humane Asso. 3% Finkelhor 1984 American Humane Asso. & & Russell National Incidence Study 20% boys 5% girls Kercher & McShane 1985 Multi-agency sample 3% Lindblad 1990 social welfare agency 5% Margolin & Craft 1989 Multi-agency sample 12.5% 92 Table 1 (con't) Percentage of Women Child Sexual Abuse Perpetrators: Perpetrators in TreatmentZEvaluation Programs (not prison) Study Date Setting % Women Abusers Faller 1987 treatment program 14% Faller 1991 treatment program 39% Groth 1979a clinical evaluation 1% Travin, Cullen & Protter 1990 treatment program 1% McCarty 1986 treatment program 4% Reports from Dav Care Center Investigations Study Date Sample % Women Abusers Faller 1988 48 abused children 2% Finkelhor & Williams 1988 national sample 21% boys 19% girls Williams & Farrell 1990 subset of a national study 38% Children Reporting Sexual Abuse by a Female Study Date Sample % Women Abusers Cupoli & Sewell 1988 emergency room 2.3% Faller 1989 clinical sample (women alone) 8% boys (women alone) 1% girls (women & men) 29% boys (women & men) 18% girls Johnson & Shrier 1987 out-patient medical clinic 46% boys Reinhart 1987 medical center 4% boys 93 Table 1 (con't) Percentage of Women Child Sexual Abuse Perpetrators; Adults Reporting Childhood Sexual Abuse by a Female Study Date Sample % Women Abusers Kendall-Tackett & Simon 1987 adults in psychotherapy 3% Self-Reports from College Students Study Date Sample % Women Abusers Condy et al. 1987 male college students 15.9% Finkelhor 1979 college students 16% boys 6% girls Fritz, Stoll & Wagner 1981 college students 60% boys 10% girls Haugaard & Emery 1989 college students 13% Landis 1956 college students 2% Risin & Koss 1987 male college students 42.7% Schultz & Jones 1983 college students 13% Self-Reports from General Population Study Date Sample % Women Abusers Finkelhor 1984 Boston sample 15% boys 0% girls Finkelhor et al.1990 national survey 17% boys 1% girls Russell 1983 random sample of women 4% Wyatt 1985 random sample of women 3% black 0% white 94 Table 2 Analysis of Variance for Aggressive Trait Rating Scale (n=144) Traits SS DF MS E p Aggressive: Target Gender 21.066 1 21.066 2.609 .109 Prime Type .111 1 .111 .014 .907 Gender X Prime 30.375 1 30.375 3.762 .054 Residual 1130.420 140 8.074 Related to Aggressive/Positive: Target Gender 15.325 1 15.325 5.867 .017 Prime Type 1.778 1 1.778 .681 .411 Gender X Prime .060 1 .060 .023 .880 Residual 365.696 140 2.612 Unrelated to Aggressive/Positive: Target Gender 19.794 1 19.794 7.511 .007 Prime Type 1.834 1 1.834 .696 .406 Gender X Prime 8.858 1 8.858 3.361 .069 Residual 368.951 140 2.635 Related to Aggressive/Negative: Target Gender 17.748 1 17.748 5.544 .020 Prime Type 7.448 1 7.448 2.327 .129 Gender X Prime 4.756 1 4.756 1.486 .225 Residual 448.193 140 3.201 Unrelated to Aggressive/Negative: Target Gender 26.090 1 26.090 9.314 .003 Prime Type 5.542 1 5.542 1.979 .162 Gender X Prime 1.273 1 1.273 .455 .501 Residual 392.153 140 2.801 95 5.5 ~fi 5.0- 9 00 l 9 o a Aggressive Trait Rating Means 15‘ Gemmu Hunk so . Male Neuteanmes Wm AggnsdwermeuwuagMM Prime Type Target gender Meptrgl Aggzessive Female M 4.34 3.34 n 35 35 Male M 4.19 5.03 n 37 37 Figure 1 96 Table 3 Analysis of Variance for the Sentence Completion Measure Prime Type SS DF MS F p Aggressive Primes (n=103): Target Gender 52.324 1 52.324 15.691 .001 Prime Type .238 1 .238 .071 .790 Gender X Prime .958 1 .958 .287 .593 Residual 333.474 100 3.335 Nurturant Primes (n=99): Target Gender 5.938 1 5.938 1.612 .207 Prime Type .260 1 .260 .071 .791 Gender X Prime 15.950 1 15.950 4.330 .040 Residual 353.637 96 3.684 Male Heterosexual Primes (n=101): Target Gender 6.265 1 6.265 1.925 .168 Prime Type 10.249 1 10.249 3.150 .079 Gender X Prime .510 1 .510 .157 .693 Residual 318.904 98 3.254 Female Heterosexual Primes (n=1o3): Target Gender 3.054 1 3.054 1.081 .301 Prime Type 4.049 1 4.049 1.433 .234 Gender X Prime 2.132 1 2.132 .755 .387 Residual 282.526 100 2.825 97 Table 4 Analysis of Variance for Nurturant Trait Rating Scale (n=144) Traits 88 DF MS F p Nurturant: Target Gender 5.025 1 5.025 .923 .338 Prime Type 10.691 1 10.691 1.963 .163 Gender X Prime 10.996 1 10.996 2.019 .158 Residual 762.423 140 5.446 Related to Nurturant/Positive: Target Gender 5.495 1 5.495 2.112 .148 Prime Type 40.047 1 40.047 15.395 .001 Gender X Prime 5.258 1 5.258 2.021 .157 Residual 364.183 140 2.601 Unrelated to Nurturant/Positive: Target Gender 4.872 1 4.872 2.400 .124 Prime Type 6.286 1 6.286 3.097 .081 Gender X Prime 6.270 1 6.270 3.089 .081 Residual 284.174 140 2.030 Related to Nurturant/Negative: Target Gender .084 1 .084 .023 .880 Prime Type .006 1 .006 .002 .969 Gender X Prime .657 1 .657 .181 .671 Residual 508.683 140 3.633 Unrelated to Nurturant/Negative: Target Gender 4.368 1 4.368 1.685 .196 Prime Type 7.467 1 7.467 2.880 .092 Gender X Prime 1.870 1 1.870 .721 .397 Residual 392.153 140 2.801 98 1.6 Sentence Stem Measure Means Gbmmu Fannie 00 ' Mme Neutral Primes Mmm NmnmmanmeuuuaGMn Prime Type Tar et e e Neutra u tu ant Female M 1.42 .78 n 26 27 Male M .19 1.15 n 27 20 Figure 2 99 Table 5 Analysis of Variance for Male Sexuality Trait Rating Scale (n=144) Traits SS DF MS E p Male Heterosexual: Target Gender 8.808 1 8.808 .820 .367 Prime Type 3.476 1 3.476 .324 .570 Gender X Prime 5.000 1 5.000 .465 .496 Residual 1514.885 141 10.744 Related to Male Heterosexuality/Positive: Target Gender 9.371 1 9.371 4.739 .031 Prime Type .024 1 .024 .012 .912 Gender X Prime .432 1 .432 .218 .641 Residual 278.802 141 1.977 Unrelated to Male Heterosexuality/Positive: Target Gender 6.745 1 6.745 2.844 .094 Prime Type 1.448 1 1.448 .611 .436 Gender X Prime 3.743 1 3.743 1.578 .211 Residual 334.354 141 2.371 Related to Male Heterosexuality/Negative: Target Gender .941 1 .941 .174 .677 Prime Type 4.890 1 4.890 .905 .343 Gender X Prime 3.708 1 3.708 .686 .409 Residual 762.265 141 5.406 Unrelated to Male Heterosexuality/Negative: Target Gender .026 1 .026 .011 .915 Prime Type 2.104 1 2.104 .923 .338 Gender X Prime .355 1 .355 .156 .694 Residual 321.450 141 2.280 100 Table 6 Analysis of Variance for Female Sexuality Trait Rating Scale (n=144) Traits ss DF MS 2 p Female Heterosexual: Target Gender 21.563 1 21.563 2.118 .148 Prime Type 7.244 1 7.244 .711 .400 Gender X Prime 1.428 1 1.428 .140 .709 Residual 1435.830 141 10.183 Related to Female Heterosexuality/Positive: Target Gender 1.820 1 1.820 .847 .359 Prime Type .469 1 .469 .218 .641 Gender X Prime 1.132 1 1.132 .527 .469 Residual 303.047 141 2.149 Unrelated to Female Heterosexuality/Positive: Target Gender 6.739 1 6.739 3.197 .076 Prime Type 3.784 1 3.784 1.795 .182 Gender X Prime .004 1 .004 .002 .967 Residual 297.242 141 2.108 Related to Female Heterosexuality/Negative: Target Gender 1.753 1 1.753 .708 .401 Prime Type .148 1 .148 .060 .807 Gender X Prime .099 1 .099 .040 .841 Residual 348.855 141 2.474 Unrelated to Female Heterosexuality/Negative: Target Gender .028 1 .028 .010 .919 Prime Type 7.237 1 7.237 2.647 .106 Gender X Prime 3.870 1 3.870 1.415 .236 Residual 385.482 141 2.734 101 Question 5: Why do you think that Donald (Donna) suggested that +3 +2 Pam (Pete) sleep in his (her) bed... child's needs mutual needs adult loves child ambivalent company for adult adult has psychological problems adult's sexual gratification Table 7 Number of Responses to Question 5 Categories (n=434) study 1 study 2 Target Gender Target Gende; Primes Female Male Female Male Score Aggressive 14 16 Male Sex 6 11 +3 Nurturant 12 7 Female Sex 17 13 Neutral ;; 1; Neutral 1g 19 39 35 41 34 Aggressive 5 3 Male Sex 6 2 +2 Nurturant 1 2 Female Sex 2 2 Neutral _g _; Neutral _5 _§ 10 7 13 10 Aggressive 2 2 Male Sex 3 1 +1 Nurturant 3 6 Female Sex 1 3 Neutral _; _; Neutral _1 ._2 6 10 5 7 Aggressive 3 0 Male Sex 1 3 0 Nurturant 1 2 Female Sex 1 4 Neutral _1 _4 Neutral _9 _; 5 6 2 10 Aggressive 8 6 Male Sex 11 1 -1 Nurturant 9 5 Female Sex 11 3 Neutral I; _§ Neutral _2 _§ 28 17 31 10 Aggressive 0 3 Male Sex 4 1 -2 Nurturant 2 0 Female Sex 0 1 Neutral _9 _; Neutral _; _Q 2 4 5 2 Aggressive 0 7 Male Sex 2 11 -3 Nurturant 3 5 Female Sex 2 6 Neutral _; _§ Neutral _Q _§ 5 20 4 23 102 Table 8 Means for Aggressive Trait Rating Scale (n=144) Traits Neutral Prime Aggressive Prime Aggressive: Gender Female 4.34 3.34 Male 4.19 5.03 Related to Aggressive/Positive: Gender Female 5.94 6.21 Male 5.33 5.51 Unrelated to Aggressive/Positive: Gender Female 6.04 6.78 Male 5.80 5.54 Related to Aggressive/Negative: Gender Female 3.36 2.53 Male 3.70 3.59 Unrelated to Aggressive/Negative: m Female 3.85 3.26 Male 4.51 4.30 103 Table 9 Means for Nurturant Trait Rating Scale (n=144) Traits Neutral Prime Nurturant Prime Nurturant: Gender Female 8.14 8.14 Male 7.22 8.31 Related to Nurturant/Positive: Gender Female 7.43 8.10 Male 6.66 8.09 Unrelated to Nurturant/Positive: Gender Female 5.41 5.41 Male 4.62 5.46 Related to Nurturant/Negative: Gender Female 4.61 4.73 Male 4.79 4.64 Unrelated to Nurturant/Negative: Gender Female 3.84 3.61 Male 4.41 3.73 104 Table 10 Means for Male Sexuality Trait Rating Scale (n=144) Traits Neutral Prime Male Sex Prime Male Heterosexual: Gender Female 5.11 5.80 Male ' 5.97 5.92 Related to Male Heterosexuality/Positive: Gender Female 6.06 6.19 Male 5.66 5.57 Unrelated to Male Heterosexuality/Positive: Gender Female 6.27 5.74 Male 5.52 5.64 Related to Male Heterosexuality/Negative: Gender Female 4.03 4.73 Male 4.51 4.57 Unrelated to Male Heterosexuality/Negative: Gender Female 3.40 3.54 Male 3.32 3.66 105 Table 11 Means for Female Sexuality Trait Rating Scale (n=144) Traits Neutral Prime Female Sex Prime Female Heterosexual: Gender Female 5.03 5.68 Male 6.00 6.25 Related to Female Heterosexuality/Positive: Gender Female 6.31 6.25 Male 5.91 6.20 Unrelated to Female Heterosexuality/Positive: Gender Female 5.66 5.33 Male 5.22 4.91 Related to Female Heterosexuality/Negative: Gender Female 3.91 3.79 Male 4.07 4.06 Unrelated to Female Heterosexuality/Negative: Gender Female 3.32 3.44 Male 3.02 3.79 106 Table 12 Means for the Sentence Completion Measure Prime Type Neutral Prime Active Prime Aggressive Primes (n=103): Gender Female 1.42 Male .19 Nurturant Primes (n=99): Gender Female 1.42 Male .19 Male Heterosexual Prunes (n-lOl): Gender Female 1.56 Male .91 Female Heterosexual Primes (n=103): Gender Female 1.56 Male .91 1.73 .10 .78 .79 .43 .90 .83 APPENDIX B Demographic Questionnaire 107 APPENDIX B Demographic Questionnaire Please answer the following questions about yourself: 1. How old are you ? 2. Are you male or female (please check one)? female male 3. Are you (please check one) Caucasian African American Native American Asian American Hispanic/Latino Other APPENDIX C Priming Stimuli 108 APPENDIX C Priming Stimuli: Aggressive/Nurturant Study Aggressive primes B. either threatens people other A. by engages fistfights in G. drunk brawls when for K. alone people insults fortune C. often acts for macho G. smashes are glass the J. others off for speaks D. over fine in cuts R. of drivers cuts off P. orders people inward around F. hits the dog at N. owns an a gun S. are loud is overbearingly R. knows where it all M. at shouts others of T. window breaks the them 8. tailgates most when driving 8. is very an unsympathetic R. can is tempered hot G. won’t take under no D. angered is easily a J. pitbull an owns a P. pushes through four crowds F. am involved is self C. insects before tortures small L. so will obey not A. accelerates under passed when N. out NRA to belongs T. animal an a kicks L. are is always rude Neutral Primes M. them crossed the street T. answered the an phone S. has have the card 8. our discussed the matter 8. the did it easily M. an saw a person G. got the a things D. bought thus book the P. away it tossed a I. thought an through it F. the test passed them C. for prepared it of N. reached it for of K. reported an it on L. date thee set the 109 [Aggressive primes distributed to students] FormzAP Gender M F (circle one) Time Began:— Age _ Time Ended:— Directions: In the following sentences the capital letter at the beginning of each line stands for the person doing the actions described in each sentence. Each sentence has been scrambled and a fourth word added. Please underline the three words that could make a complete sentence. For example: P. dog the an found = P. found the dog. You would underline the words ”found", "the", and ”dog”. In other words, P. d_og ;h_e an found. Work as quickly as you can without making any mistakes. Please time how long this takes you to complete this exercise. B. either threatens people other A. by engages frstfrghts in M. them crossed the street T. animal an a kicks K. alone people insults fortune T. answered the an phone C. often acts for macho G. smashes are glass the S. has have the card J. others off for speaks D. over line in cuts K. reported an it on R of drivers cuts off P. orders people inward around L. date thee set the F. hits the dog at N. owns an a gun 8. our discussed the matter 8. are loud is overbearingly 110 R. knows where it all B. the did it easily M. at shouts others of T. window breaks the them M. an saw a person S. tailgates most when driving B. is very an unsympathetic G. got the a things R. can is tempered hot G. won’t take under no P. away it tossed a D. angered is easily a J. pitbull an owns a F. the test passed them P. pushes through four crowds F. am involved is self D. bought thus book the C. insects before tortures small L. so will obey not J. thought an through it A. accelerates under passed when N. out NRA to belongs C. for prepared it of G. drunk brawls when for L. are is always rude N. reached it for of 111 Priming Stimuli: Aggressive/Nurturant Study Nurturant Primes . has close relationships . worries about feelings . makes loving home . is quite gentle . is very understanding . is a nurse . is a socialworker . holds crying children . buys children's clothes 10. is quite patient 11. is a caregiver I2. is always giving 13. provides emotional support 14. concerned about others 15. caretakes during sickness l6. participates in PTA 17. plants flower gardens 18. waters the plants 19. cleans the house 20. sends care packages 21. provides never-ending support 22. nurses the sick 23. consoles a friend 24. puts others first 25. notices children’s problems 26. senses people’s feelings 27. is very kind 28. does the laundry 29. makes selfless sacrifices 30. verbally rewards children ~0Nfl0~khbWN—n Neutral Primes M. them crossed the street T. answered the an phone 8. has have the card B. our discussed the matter B. the did it easily M. an saw a person G. got the a things D. bought thus book the P. away it tossed a J. thought an through it F. the test passed them C. for prepared it of N. reached it for of K. reported an it on L. date thee set the A. has relationships close so B. feelings worries out about G. makes loving is home K. is either quite gentle C. understanding is very inward G. is most nurse a J. where is socialworker a D. holds under crying children R clothes children’s buys before P. can is quite patient F. is a am caregiver N. is giving always are S. a emotional provides support R. concemed under others about M. sickness caretakes during of T. participates by in PTA S. plants flower are gardens B. the waters plants off R. the house a cleans G. packages care sends over D. support provides never-ending an J. of the nurses sick P. friend consoles a four F. a first puts others C. are problems children’s notices L. people’s them feelings senses A. is an kind very N. am laundry the docs T. fortune sacrifices selfless makes L. verbally rewards is children 112 [Nurturant primes distributed to students] FormzNP Gender M F (circle one) Time Began:— Age _ Time Ended:— Directions: In the following sentences the capital letter at the beginning of each line stands for the person doing the actions described in each sentence. Each sentence has been scrambled and a fourth word added. Please underline the three words that could make a complete sentence. For example: P. dog the an found = P. found the dog. You would underline the words ”found", “the“, and 'dog". In other words, P. d_og .th_g an found. Work as quickly as you can without making any mistakes. Please time how long this takes you to complete this exercise. A. has relationships close so R. the house a cleans M. them crossed the street N. is giving always are B. feelings worries out about T. answered the an phone R. clothes children’s buys before G. makes loving is home S. has have the card K. is either quite gentle B. the waters plants off M. an saw a person C. understanding is very inward G. is most nurse a B. our discussed the matter D. holds under crying children P. can is quite patient B. the did it easily T. participates by in PTA 113 F. is a am caregiver G. got the a things P. friend consoles a four S. a emotional provides support D. bought thus book the R. concerned under others about M. sickness caretakes during of J. thought an through it S. plants flower are gardens J. where “m socialworker a P. away it tossed a G. packages care sends over D. support provides never-ending an L. date thee set the J. of the nurses sick F. a first puts others C. are problems children’s notices F. the test passed them A. is an kind very N. am laundry the docs T. fortune sacrifices selfless makes N. reached it for of L. verbally rewards is chilan C. for prepared it of L. people’s them feelings senses K. reported an it on Priming Stimuli: Sexuality Study Male Sexual Primes . cheats on people . will abuse people . says disgusting things . initiates sexual advances needs no foreplay . tears off clothes . frequents strip clubs . possessive of spouse . ogles at bars 10. likes rough sex 11. brags about sex 12. is outwardly aggressive 13. has the authority 14. afraid of emotions 15. isn’t easily attached 16. looks at Playboy l7. interested in conquering 18. brags about conquests 19. can be aggressive 20. wants no romance 21. likes to masturbate 22. wants more sex 23. really wants power 24. is more aggressive 25. is sexually dominant 26. says vulgar things 27. only wants sex 28. fears close relationships 29. will show off 30. likes one-night stands sooeqasytawm... Neutral Primes M. them crossed the street T. answered the an phone S. has have the card B. our discussed the matter B. the did it easily M. an saw a person G. got the a things D. bought thus book the P. away it tossed a J. thought an through it F. the test passed them C. for prepared it of N. reached it for of K. reported an it on L. date thee set the 114 B. either cheats people on A. by will people abuse G. things says disgusting for K. advances sexual initiates fortune C. needs no for foreplay G. tears are clothes off 1. clubs off stn'p frequents D. are spouse of possessive R. of bars ogles at F. likes a sex rough P. brags about inward sex N. is an outwardly aggressive S. are authority has the R afraid where of emotions M. easily isn’t attached of T. Playboy looks at they S. interested may in conquering B. brags about an conquests R. is can aggressive be G. wants no under romance D. masturbate likes to a J. sex an wants more P. really wants four power F. are aggressive is more C. dominant before is sexually L. am things says vulgar A. only under sex wants N. out relationships close fears T. off show a will L. over stands one-night likes 115 [Male heterosexual primes distributed to students] ForrnzMSP Gender M F (circle one) Time Began:— Age _ Time Ended:— Directions: In the following sentences the capital letter at the beginning of each line stands for the person doing the actions described in each sentence. Each sentence has been scrambled and a fourth word added. Please underline the three words that could make a complete sentence. For example: P. dog the an found = P. found the dog. You would underline the words "found", “the", and ”dog”. In other words, P. d_og _th_e an found. Work as quickly as you can without making any mistakes. Please time how long this takes you to complete this exercise. D. are Spouse of possessive B. brags about an conquests M. them crossed the street B. either cheats people on T. answered the an phone A. by will people abuse G. things says disgusting for S. has have the card advances sexual initiates fortune .7“ . reported an it on 2 7S . is an outwardly aggressive . our discussed the matter a C. needs no for foreplay R. afraid where of emotions B. the did it easily r . over stands one-night likes . bought thus book the . tears are clothes off 0 D . reached it for of Z 116 G. wants no under romance A. only under sex wants G. got the a things R of bars ogles at J. thought an through it P. brags about inward sex S. are authority has the M. easily isn’t attached of J. clubs off strip frequents P. away it tossed a T. Playboy looks at they S. interested may in conquering L. date thee set the R is can aggressive be D. masturbate likes to a F. the test passed them J. sex an wants more P. really wants four power N. out relationships close fears F. are aggressive is more T. off show a will C. dominant before is sexually M. an saw a person L. am things says vulgar C. for prepared it of F. likes a sex rough Priming Stimuli: Sexualig Study Female Sexual Primes . will lower eyes wants door opened . loves to flirt is somewhat weaker is somewhat submissive . must feel beautiful . timid about body . dresses for others . wears provocative clothing 10. seeks older partners 11. must be thin 12. seeks taller partners l3. finds sex boring l4. wears tight clothes 15. blinded by feelings 16. should look attractive l7. wears low-cut tops 18. won’t initiate sex 19. can walk seductively 20. is sexually inhibited 21. isn’t sexually aggressive 22. focuses on appearance 23. experiences domestic abuse 24. has mood changes 25. influenced by others 26. is easily attached 27. is physically clingy 28. wears revealing clothes 29. plays with hair 30. wants ”fairytale” love coexroytgswp— Neutral Primes M. them crossed the street T. answered the an phone S. has have the card B. our discussed the matter B. the did it easily M. an saw a person G. got the a things D. bought thus book the P. away it tossed a J. thought an through it F. the test passed them C. for prepared it of N. reached it for of K. reported an it on L. date thee set the 117 A. will eyes lower so B. opened wants out door G. loves to for flirt K. is either somewhat weaker C. submissive is somewhat most G. must inward beautiful feel J. where timid body about D. dresses under for others R clothing provocative wears before P. can seeks older partners F. must be am thin N. seeks partners taller are S. under sex finds boring R wears a clothes tight M. feelings blinded by of T. should by look attractive S. wears are low-cut tops B. initiate won’t sex off R walk seductively a can G. inhibited sexually is may D. aggressive isn’t sexually an J. of on focuses appearance P. abuse experiences domestic a F. four changes has mood C. are others by influenced L. easily them attached is A. is an clingy physically N. am clothes revealing wears T. fortune hair with plays L. wants "fairytale” is love 119 R. wears a clothes tight G. got the a things M. feelings blinded by of T. should by look attractive J. thought an through it 8. wears are low-cut tops P. away it tossed a B. initiate won’t sex off F. the test passed them R walk seductively a can C. for prepared it of D. aggressive isn’t sexually an J. of on focuses appearance N. reached it for of R clothing provocative wears before P. abuse experiences domestic a F. four changes has mood C. are others by influenced L. easily them attached is A. is an clingy physically L. date thee set the G. inhibited sexually is may T. fortune hair with plays L. wants ”fairytale" is love M. an saw a person N. am clothes revealing wears 120 Priming Stimuli: Neutral Condition Neutral Stimuli M. them crossed the street T. lifted them bag the L. of the job did A. wrote where paper the T. answered the an phone B. for a hoped it 8. has have the card G. the drove car are K. dog the walked fortune B. our discussed the matter C. reporter is a an G. has a inward beeper J. of door closed the B. the did it easily D. guitar over wanted the M. an saw a person R the doubted off reason P. are idea the liked G. got the a things F. is enjoyed meal a N. heard for song a 8. saw a either movie D. bought thus book the R. read book by the P. away it tossed a M. again are will practice T. a last arrived at J. thought an through it S. bought a ticket under B. can went class to R. is out student a G. the takes train am F. the test passed them D. often a comes over C. for prepared it of J. so shirt has a N. reached it for of P. an ball the threw F. you spoke most of C. under the sound heard L. marked before line the K. reported an it on A. saw an the train N. locked four window the L. date thee set the 121 [Neutral primes distributed to students] Form:CP Gender M F (circle one) Time Began:— Age _ Time Ended:— Directions: In the following sentences the capital letter at the beginning of each line stands for the person doing the actions described in each sentence. Each sentence has been scrambled and a fourth word added. Please underline the three words that could make a complete sentence. For example: P. dog the an found = P. found the dog. You would underline the words "found“, “the", and "dog". In other words, P. deg £13 an found. Work as quickly as you can without making any mistakes. Please time how long this takes you to complete this exercise. M. them crossed the street T. lifted them bag the L. of the job did A. wrote where paper the T. answered the an phone B. for a hoped it S. hm have the card G. the drove car are K. dog the walked fortune B. our discussed the matter C. reporter is a an G. hm a inward beeper J. of door closed the B. the did it emily D. guitar over wanted the M. an saw a person R the doubted off reason P. are idea the liked G. got the a things 122 F. is enjoyed meal a N. heard for song a S. saw a either movie D. bought thus book the R read book by the P. away it tossed a M. again are will practice T. a last arrived at J. thought an through it S. bought a ticket under B. can went class to R is out student a G. the takes train am F. the test pmsed them D. often a comes over C. for prepared it of J. so shirt hm a N. reached it for of P. an ball the threw F. you spoke most of C. under the sound heard I" . marked before line the . reported an it on > 7< . saw an the train 2 . locked four window the L. date thee set the APPENDIX 1) Pilot Study 123 APPENDIX D Pilot Study Development of the Priming Stimuli Twenty-two pilot participants (6 men and 16 women) generated sentences of behavior characteristic of nurturance in women, of male heterosexuality, and of female heterosexuality. From this pool of sentences, separate lists of nurturant, male sexual behavior and female sexual behavior were compiled. Each of the behavior descriptions was simplified into a three-word phrase. A separate pilot sample of 44 participants (21 men and 23 women) each rated half of the behaviors on one of two dimensions: negativity (l--not at all negative to 7--extremely negative) and gender typicality (1--completely female to 7--completely male) (cf. Banaji et al. , 1993). Nurturant items were selected for use as primes when: l) the mean ratings for male and female raters did not differ by more than one point for negativity and gender ratings, 2) nurturant items were rated less than or equal to a mean of 3.0 on negativity, and 3) nurturant items had a mean of 3.1 or lower on gender typicality (i.e. they were typical of females) (cf. Banaji et al., 1993). Male sexual items were selected for use as primes when: 1) the mean ratings for male and female raters did not differ by more than one point for gender, and by more than two points for negativity, 2) male sexual items were rated greater than or equal to a mean of 3.0 on negativity, and 3) male sexual items had a mean of 4.9 or higher on gender typicality (i.e. they were typical of males) (cf. Banaji et al., 1993). Female sexual items were selected for use as primes when: l) the mean ratings for male and female raters did not differ by more than one point for gender, and by more than two points for negativity, 2) female sexual items were rated greater than or equal to a mean of 3.0 on negativity, and 3) female sexual items had a mean of 3.3 or lower on gender typicality (i.e. they were typical of females) (cf. Banaji et al., 1993). Development of the Target Paragr_aphs The initial vignette was modified after piloting work demonstrated that there was little variability; the majority of subjects (N=l8) saw the vignette as abusive. Subjects suggested that the age of the child be lowered, and that more aggressive behaviors be included. After these changes, a second group of pilot subjects (N =4) still found the story abusive. They suggested that the adult change into pajamas before getting into bed with the child. A third group of subjects (N =26) suggested that various suggestive adjectives (e.g. nice, warm, carefully, slowly) used in the vignette be removed so that the story became more ambiguous. The fourth group of subjects (N= 13) suggested that the ambiguous incidents be scattered throughout the vignette, and that the word "body" be removed as it was too ”alerting”. 124 In the final version of the story, the child’s age was lowered from 8 to 6, an additional sexually ambiguous incident was added to the beginning of the vignette, and one neutral sentence was deleted as the vignette was getting too long and detailed. Only a small minority (15%) of the subjects (N =22) found this vignette abusive. Development of the Sentence Stem Measure This measure was designed so that participants could describe how they interpreted the behaviors in the target vignette. Using a checklist or other type of scale imposes the experimenter’s construction of a situation onto the participant; by leaving the question stems Open ended this scale measured how participants interpreted ambiguous behavior. This scale was pretested along with the vignette and a number of subtle changes in the wording of the directions and phrasing of the sentence stems were made so that enough structure was provided to insure that participants would explain the behavior of the protagonist. An additional question, (#9), was added after it became apparent that some participants were not expressing their emotional reaction to the story using the existing stems. This measure consisted of 8 open ended questions and I forced choice question. The first two questions were neutral questions. Of the remaining six questions, three, (numbers 3, 5, and 7), asked participants to explain sexually ambiguous behaviors, and two questions, (numbers 4 and 6), could be responded to as neutral questions or as questions about sexually ambiguous behaviors. Question number 8 was a forced choice question, and number 9 asked participants what feeling(s) they had after reading the story. The first two questions were neutral and were not intended to be included in the scale; of the remaining 6 open-ended questions, (numbers 3, 4, 5, 6, 7, and 9), only 3, (numbers 3, 5, and 9), yielded useful data. Unfortunately, most participants answered questions 4, 6, and 7, exactly as they were presented in the story thus yielding very little variability. For example, on question 7 ”Donald (Donna) got in bed next to Pam (Pete). . . " out of a subsample of 168 subjects, only 7 did not respond with "and held her (him) tight, " the exact statement in the story. This left only three questions that provided useful data from which to construct a scale. This measure was coded using the dimension appropriate/ inappropriate to code behaviors, and the dimension good/bad to code feelings. All three questions were given a 1 for a response that was deemed appropriate (or a good feeling), a 0 for an ambivalent response, (i.e. those responses that contained both an appropriate and an inappropriate reason), and a -l for a response that gave an inappropriate reason (or listed a bad feeling). Scores on the sentence stem measure ranged from +3 to -3. This scale had an internal consistency reliability (Cronbach’s Alpha) of .60. APPENDIX E Target Paragraphs 125 APPENDIX E Male Target Paragr_aph Donald picked up 6 year old Pam from school and gave her a quick kiss "hello". They went to the park and sat down next to each other on a bench. AS Pam told Donald about her day at school, he leaned toward her, nodding enthusiastically and stroked her shoulder. Pam saw two of her friends, got up and went to play on the swings with them while Donald sat on the bench and read the newspaper. Once Donald and Pam got home they ordered Chinese food for dinner. After they had finished eating dinner, Pam helped Donald clear the table and put away the left-overs. She did her homework and played computer games until it was her bed time. Donald told Pam that she needed a bath that evening, so he ran a warm bubble bath for her. He supervised Pam getting undressed and washing in the bathtub. Donald helped Pam wash with her favorite bubble gum scented soap, he also washed her hair. Afier the bath, he helped Pam dry herself and put her pajamas on. He suggested that she sleep in his bed that evening. Pam got into bed, and Donald hugged her and kissed her goodnight. Donald spent the rest of the evening listening to some CD’S, and watching television. When he grew tired, he went into his bedroom and undressed. Afier changing into his pajamas, Donald got into bed next to Pam and pulled her towards him. 126 Female Target Paraggaph Donna picked up 6 year old Pete from school and gave him a quick kiss "hello". They went to the park and sat down next to each other on a bench. As Pete told Donna about his day at school, she leaned toward him, nodding enthusiastically and stroked his shoulder. Pete saw two of his friends, got up and went to play on the swings with them while Donna sat on the bench and read the newspaper. Once Donna and Pete got home they ordered Chinese food for dinner. After they had finished eating dinner, Pete helped Donna clear the table and put away the left-overs. He did his homework and played computer games until it was his bed time. Donna told Pete that he needed a bath that evening, so she ran a warm bubble bath for him. She supervised Pete getting undressed and washing in the bathtub. Donna helped Pete wash with his favorite bubble gum scented soap, she also washed his hair. After the bath, she helped Pete dry himself and put his pajamas on. She suggested that he sleep in her bed that evening. Pete got into bed, and Donna hugged him and kissed him goodnight. Donna spent the rest of the evening listening to some CD’S, and watching television. When she grew tired, she went into her bedroom and undressed. After changing into her pajamas, Donna got into bed next to Pete and pulled him towards her. 127 Story Sentence Content Analysis 1. Donald picked up 6 year old Pam from school, (neutral) 2. and gave her a quick kiss "hello". (sexually aggressive or nurturant) 3. They went to the park and sat down next to each other on a bench. (neutral) 4. As Pam told Donald about her day at school, he leaned toward her, nodded enthusiastically and stroked her shoulder. (sexually aggressive or nurturant) 5. Pam saw two of her friends, got up and went to play on the swings with them while Donald sat on a bench and read the newspaper. (neutral) 6. Once Donald and Pam got home they ordered Chinese food for dinner. (neutral) 7. After they had fmished eating dinner, Pam helped Donald clear the table and put away the left-overs. (neutral) 8. She did her homework and played computer games until it was her bed time. (neutral) 9. Donald told Pam that she needed a bath that evening, so he ran a warm bubble bath for her. (sexually aggressive or nurturant) 10. He supervised Pam getting undressed and while washing in the bathtub. (sexually aggressive or nurturant) 11. Donald helped Pam wash with her favorite bubble gum scented soap, he also washed her hair. (sexually aggressive or nurturant) 12. After the bath, he helped Pam dry herself and put her pajamas on. (sexually aggressive or nurturant) 13. He suggested that she sleep in his bed that evening. (sexually aggressive or nurturant) 1 2 8 14. Pam got into bed, and Donald hugged her and kissed her goodnight. (sexually aggressive or nurturant) 15. Donald spent the rest of the evening listening to some CD’S, and watching television. (neutral) 16. When he grew tired, he went into his bedroom and undressed. (neutral) 17. After changing into his pajamas, Donald got into bed next to Pam and pulled her towards him. (sexually aggressive or nurturant) Neutral statements - 8 Sexually aggressive or nurturant statements - 9 APPENDIX F Filler Task 129 APPENDIX F "7" Please read the following sentences and cross out all the letter e s . \OOOqakh-hUJNu—r . I have a good appetite. . I wake up fresh and rested most mornings. . I think I would like the work of a librarian. . I am easily awakened by noise. . My hands and feet are usually warm enough. . My daily life is full of things that keep me interested. . I am about as able to work as I ever was. . There seems to be a lump in my throat much of the time. . A person should try to understand his dreams and be guided by or take warning from them. 10. ll. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. I work under a great deal of tension. I have diarrhea once a month or more. Once in a while I think of things too bad to talk about. I am sure I get a raw deal from life. I used to have imaginary companions. I am very seldom troubled by constipation. Usually I would prefer to work with women. At times I have fits of laughing and crying that I cannot control. I am troubled by attacks of nausea and vomiting. No one seems to understand me. I would like to be a singer. I feel sure that there is one true religion. I am bothered by acid stomach several times a week. I have nightmares every few nights. I find it hard to keep my mind on a task or job. I have had very peculiar and strange experiences. I have a cough most of the time. If people had not had it in for me I would have been much more successful. I seldom worry about my health. Most any time I would rather Sit and daydream than to do anything else. I have had periods of days, weeks or months when I couldn’t take care of things because I couldn’t "get going." 31. My family does not like the work I have chosen (or the work I intend to choose for my life work). 32. 33. My sleep is fitful and disturbed. Much of the time my head seems to hurt all over. 34. I do not always tell the truth. 130 35. My judgment is better than it ever was. 36. Once a week or oftener I feel suddenly hot all over, without apparent cause. 37. It makes me nervous to have to wait. 38. My soul sometimes leaves my body. 39. I am in just as good physical health as most of my friends. 40. I prefer to pass by school friends, or people I know but have not seen for a long time, unless they Speak to me first. 41. I am liked by most people who know me. 42. I am almost never bothered by pains over the heart or in my chest. 43. I am a good mixer. ‘ 44. Everything is turning out just like the prophets of the Bible said it would. 45. I do not read every editorial in the newspaper every day. 46. I have not lived the right kind of life. 47. Parts of my body often have feelings like burning, tingling, crawling, or like "going to sleep." 48. I have had periods in which I lost sleep over worry. 49. I see things or animals of people around me that others do not see. 50. I wish that I could be as happy as others seen to be. 51. I hardly ever feel pain in the back of my neck. 52. I think a great many people exaggerate their misfortunes in order to gain the sympathy and help of others. 53. I am troubled by discomfort in the pit of my stomach every few days or oftener. 54. I am an important person. 55. Most of the time I feel blue. 56. I like poetry. 57. My feelings are not easily hurt. 58. I think I would like the kind of work a forest ranger does. 59. I am easily downed in an argument. 60. Any man who is able and willing to work hard has a good chance of succeeding. 61. These days I find it hard not to give up hope of amounting to something. 62. I am certainly lacking in self-confidence. 63. I would like to be a florist. 64. I usually feel that life is worth while. 65. Once in a while I put off until tomorrow what I ought to do today. 66. I do not mind being made fun of. 67. I go to church almost every week. 68. I have very few quarrels with members of my family. 69. I believe in the second coming of Christ. 70. I have met problems so full of possibilities that I have been unable to make up my mind about them. 71. My hardest battles are with myself. 72. I have little or no trouble with my muscles twitching or jumping. 73. I don’t seem to care what happens to me. 74. I am happy most of the time. 131 75. There seems to be a fullness in my head or nose most of the time. 76. Someone has it in for me. 77. I have never done anything dangerous for the thrill of it. 78. I am apt to pass up something that I want because others feel that I am not going about it the right way. 79. Often I feel as if there were a tight band about my head. 80. I believe in a life hereafter. 81. I enjoy a race or game better when I bet on it. 82. Most people are honest chiefly through fear of being caught. 83. My speech is the same as always (not faster or slower, or slurring; no hoarseness). 84. My table manners are not quite as good at home as when I am out in company. 85. I believe that I am being plotted against. 86. I seem to be about as capable and smart as most others around me. 87. I believe that I am being followed. 88. Most people will use somewhat unfair means to gain profit or an advantage rather than lose it. 89. I have a great deal of stomach trouble. 90. I like dramatics. 91. I know who is responsible for most of my troubles. 92. The sight of blood neither frightens me nor makes me sick. 93. I have never vomited blood or coughed up blood. 94. I like collecting flowers or growing house plants. 95. At times my thoughts have raced ahead faster than I could Speak them. 96. If I could get into a movie without paying and be sure I was not seen I would probably do it. 97. I commonly wonder what hidden reason another person may have for doing something nice for me. 98. I am bothered by people outside, on the streetcars, in stores, etc., watching me. 99. Criticism or scolding hurts me terribly. 100. My conduct is largely controlled by the customs of those about me. 101. I certainly feel useless at times. 102. When I was a child, I belonged to a crowd that tried to stick together through thick and thin. 103. I have the wanderlust and am never happy unless I am roaming or traveling about. 104. I have often lost out on things because I couldn’t make up my mind soon enough. 105. I used to keep a diary. 106. Someone has been trying to poison me. 107. Most nights I go to sleep without thoughts or ideas bothering me. 108. During the past few years I have been well most of the time. 109. I have never had a fit or convulsion. 110. I am neither gaining nor losing weight. 111. I have had periods in which I carried on activities without knowing later what I had been doing. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 132 I feel that I have often been punished without cause. I cry easily. I cannot understand what I read as well as I used to. I have never felt better in my life than I do now. The top of my head sometimes feels tender. I very seldom have spells of the blues. I do not tire quickly. I like to study and read about things that I am working at. I like to know some important people because it makes me feel important. I am afraid when I look down from a high place. There is something wrong with my mind. I am not afraid to handle money. What others think of me does not bother me. It makes me uncomfortable to put on a stunt at a party even when others are doing the same sort of things. 126 127 128 129 130 131. 132. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155 I frequently have to fight against showing that I am bashful. I liked school. I have never had a fainting Spell. I seldom or never have dizzy spells. I do not have a great fear of snakes. My mother was a good woman. My memory seems to be all right. I find it hard to make talk when I meet new people. I am afraid of losing my mind. I am against giving money to beggars. My hearing is apparently as good as that of most people. I frequently notice my hand shakes when I try to do something. I can read a long while without tiring my eyes. Sometimes, when embarrassed, I break out in a sweat which annoys me greatly. I do not have spells of hay fever or asthma. I do not like everyone I know. I like to visit places where I have never been before. I daydream very little. I wish I were not so shy. If I were a reporter I would very much like to report news of the theater. I would like to be a journalist. I am very religious (more than most people). I enjoy many different kinds of play and recreation. I believe that my sins are unpardonable. Everything tastes the same. I can sleep during the day but not at night. I could be happy living all alone in a cabin in the woods or mountains. In walking I am very careful to step over sidewalk cracks. I have never had any breaking out on my skin that has worried me. 133 158. I have used alcohol excessively. 159. I believe that a person should never taste an alcoholic drink. 160. I think I would like the work of a building contractor. 161. I feel unable to tell anyone all about myself. 162. I like science. 163. It is not hard for me to ask help from my friends even though I cannot return the favor. 164. I have used alcohol moderately (or not at all). 165. I gossip a little at times. 166. I used to like hopscotch. 167. I have been told that I walk during sleep. 168. At times I feel that I can make up my mind with unusually great ease. 169. I should like to belong to several clubs or lodges. 170. I have several times had a change of heart about my life work. APPENDIX G Sentence Stems 134 APPENDIX G Male Story Sentence Stems Based on the story that you read earlier, please complete the following sentences. For the sentences that the story does not explicitly answer, make-up answers as you understand the Story: 1. After Donald picked up Pam from school, they went to 2. Donald and Pam ate dinner and then 3. Donald helped Pam bathe because 4. After the bath, Donald and Pam 5. Why do you think Donald suggested that Pam sleep in his bed 6. Pam went to sleep and Donald 7. Donald got in bed next to Pam 135 8. In your opinion is Donald Pam’s (check one) i _father? _brother? _babysitter’? _uncle? _neighbor'? _other ? (please specify) 9. What feeling(s) did you have after reading this story? 136 APPENDIX G Female Story Sentence Stems Based on the story that you read earlier. please complete the following sentences. For the sentences that the story does not explicitly answer, make-up answers as you understand the story: 1. After Donna picked up Pete from school, they went to 2. Donna and Pete ate dinner and then 3. Donna helped Pete bathe because 4. After the bath, Donna and Pete 5. Why do you think Donna suggested that Pete sleep in her bed 6. Pete went to Sleep and Donna 7. Donna got in bed next to Pete 137 8. In your opinion is Donna Pete’s (check one) _mother? _sister? _babysitter? _aunt? _neighbor? _other ? (please specify) 9. What feeling(s) did you have after reading this story? APPENDIX H Scoring Manual for Sentence Stems 138 APPENDIX H Scoring Manual for Sentence Stems Question 3 D. helped P. bathe because... scored from + l to -1 +1: 0.1 = 0.2 = appropriate vague ambivalent Recoded as 0. -l = inappropriate +1= appropriate-In these respomes appropriate reams for helping a child bathearegiven;snehmage,safetyconeansorb’santhwityuacnretakais examples: age of child child was dirty/needed a bath child needs supervision child needs help/assistance to insure child’s safety D is the mother/ father D loves P, wants to be nice child wants to be helped D felt P needed to D usually did D was the one who told P to take a bath 139 0.l=vague In these responses mbjects do not articulate an appropriate or an inappropriate resson for helping P. bathe; these responses are generally vague and notenoughinformafiontodeterminewhethertheresponseisappmpriateor inappropriate. examples: D wanted to D needed to D had nothing else better to do D liked to watch P gd because P is a young child D was going to clean the tub as well D liked to D wanted to...hmm... to watch P 0.2 = ambivalent-In these responses the subject articulates appropriate reasons (suchastheageofthechild), and inappropriatereasons(snchmseeingthechild nude) for helping P. bathe. Subjects do not state that the action either appropriate or inappropriate; it could be either. examples: P is young & needs help 9; D wants to see P naked. didn’t want P to drown pr; D’s a pervert -l = inappropriate-In these responses the subject articulates an inappropriate ressonforanadufltohdpachfldbathqsuchastheadnlt’ssexnalpleasme. examples: D is lonely & can’t differ. between a child & an adult D is a weirdo D’s motives stink D is a child molester/ pervert! pedophile D was in need of sexual gratification D likes P in an unhealthy way D wants to see P nude 140 Question 5 Why do you think D suggested that P sleep in his/her bed scored from + 3 to -3 Rescored for the Sentence Stem Measure. Scored along a continuum of needs; from child’s needs (+3) to adult’s sexual needs (-3). If a subject gives a response with two scorable positive or negative statements, score the extreme number. For example: D is lonely (= -1 adult needs for company) and possibly even a pervert (-3 adult sexual needs) is scored as -3. If a response contains positive and negative statements and is ambiguous, it is scored as 0.2. For example: One or the other might have needed comforting; (+2) they may have only had one bed; (+2) sexual reasons (-3) is scored as 0.2. Sentence Stem + 1 +1 +1 Descriptive Data +3 = meeting child’s needs + 2 = mutual needs + l = adult loves/misses the child 0. l = uncodable/vague 0.2 = ambivalent -l = meeting adult needs for company -2 = adult has psychological problems -3 = meeting adult’s sexual needs 141 +3 = Intheseresponsesit'nspeeifieallystatedthatthechildwouldfeelmore comfortable sleeping in the adult’s bed for emotional or phyu'cal reasons. The emphasis k on the child’s comfort. examples: child was afraid of something P is scared or hasn’t been sleeping well P has been having nightmares so P would feel safe D had gotten divorced and P was upset P had witnessed a murder recently and didn’t want to sleep alone to protect P P enjoys being close to D when P sleeps P is young and in an unfamiliar place because P did not feel well in case there are creepy, crawly monsters under P’s bed because P is young D wants P to feel safe D wanted to comfort P P is afraid of the dark D wanted P to know that D loved P Because P would feel special to be nice P would feel more comfortable P had no where else to sleep didn’t want P to sleep on the couch or floor because D didn’t have a better place for P because P will sleep better so that P would not be alone 142 +2 = Intheseresponsesboththechildandtheadult’sneedsforsecurity (safety), love or comfort are met by having the child Sleep in the adult’s bed (or it’simpliedthatbothoftheirneedsarebeingmet). A-lresponsecanbemoved uptoa+2whenarefereneetothechild’sneedsareincluded. +1 examples: D’s the parent, misses P and P had nightmares the night before D is P’s parent and it may have been thundering D wanted to be with P and it was supposed to storm because P always ends up in D’s bed anyway because it was a stormy night if P needed anything D would be right there and D wanted P close in order to make P and D feel safe because it was more comfortable and D didn’t want P to be alone to keep each other company so P didn’t get scared, or D was lonely P enjoys sleeping there and D wanted the company because it was bigger and D wanted P nearby Intheseresponsestheadultlovesorueetheehild. examples: because D loves P because D loves P so much D missed P and recently got custody of P D missed P 143 0.1 = uncodable-Thesestatementsareuncodable. 'I‘heydonotfitintoanyofthe categofim;generaflybecausetheyarevagudyworded,andmtfldbehnapreted aseitherpositiveornegative. examples: with her? I don’t know/ not enough information after the bath? it could be for a number of reasons heck if I know they were lovers they were the only ones home because Mom/ Dad was not there maybe wife/husband was out of town 0.2 = ambivalent-Mrmponsesinclndebothastatanmtthatconldhecoded positivelyornegatively. 'I‘heseanswersareambivalent. examples: PisscaredofthedarkpgDisasicko It was cold out g D is a pedophile. one may have needed comforting, 1 bed pr; sexual reasons D loves P and is lonely Maybe P has nightmares and often wakes D up in the middle of the night (I hope). 144 -1 = Intheseresponsestheadult’sdesiretohavethechfldphysicaflydose(often becausetheadultislonely) isgivenasthereason forthechildtosleepinthe adult’s bed. The emphasis is on the adult; only adult needs are mentioned, no consideration is given for the child. When both adult and child’s needs are stated it is scored as a +2. examples: to keep adult company D didn’t want to sleep alone D needs someone to sleep with so their day of togetherness wouldn’t end for D D feels lonely (maybe Spouse is out of town) D liked the company D is scared about something D wanted to feel close to P because D was sad D wanted P close to him/her -2 = Intheseresponsestheadultisidmfifiedmhavingpsychologicalpmbluns. examples: D has mental problems because D is a very sick person D wanted P’s company and was sick minded D’s sick in the head because he was "sick" 145 -3 = Intheseresponsestheadult’ssexualneedsaredesrlystatedasthereason forsuggestingthatthechildsleepintheadultsbed. examples: it seemed perverted wanted to "sleep" next to child D has a sick mind and is molesting P D is a pedophile/ wanted to molest P D is lonely and possibly even a pervert psychopathic child molester incestuous thoughts D is sick and perverted so that D could feel on P D has sexual feelings toward P so D could molest P 146 Question 9 What feeling(s) did you have after reading this story? Scored from +1 to -1 + l = good 0.1 = unscorable, vague, or no feelings 0.2 = confused or ambivalent feelings Recoded as 0. -1 = bad +1 = good-Theseareresponsesinwhichthesubjectreportspoa‘fivefeelingsor examples: D is a nurturing type of person D loves P and takes care of P maternal, close parent/child relationship D loves P, I felt all warm inside they have a strong bond a caring feeling P needs TLC and gets tender loving D is P’s older sibling and likes to take care of P D is taking care of P alone and showing P lots of love D is a nice sibling helping out P a love bond between parent and child a nice, happy, loving story D loves his/her child a lot a parent wanting to be emotionally close to a child a caring parent a strong caring family and D is careful with P love and caring D cares greatly about P a sense of love for the child felt good because D and P get along so well a good parent who took care of P real well they loved and trusted each other I felt warm and fuzzy D is a caregiver to P a cute story describing a typical day pleasant they were a very close family 147 0.1 = vague- 'I‘hesubjectdoesnotstateanyfeelingsortheresponseisvague, strangeornnscorable. examples: no feelings: none, the story was pointless none not really no feelings I felt like going to sleep wondering when I would have to cross out all the e’s typical situation in a parents day a story about a parent and child didn’t know if D was a parent or someone else related to P at first the last line made no sense, but later it did D’s spouse is not at home or D is a single parent wondered where the other parent was I thought the story was told rather intimately none really, just that D seems to be protective a parent and child living without a spouse D has a lot of feelings for P that it was a divorced family many explanations but not enough info strange: sexual feelings between the two it got me a little excited 148 0.2 = ambivalent- The subject is confused by the story, or perceives the story ambivalently; often because they are not sure of the relationship between D. and P (the relationship could be interpreted positively or negatively). examples: confused: confusion mixed, I’m left in a mystery don’t know how to feel, kinky inferences but no concrete molestation positive or negative: normal night or the beginnings of molestation sexually hitting on P or trying to be loving D loves P a lot but maybe D is too affectionate don’t know if gestures were friendly or sexually abusive parent reference: D is a loving parent or a child molester confused as to whether D was a pervert or just being a parent frustrated, I want to know if D is P’s parent for sure if D was the parent, D shouldn’t be so close to P seemed like a child molester but if D is a parent than its ok it was strange if D was the parent uneasiness as to who D was but P was comfortable enough unsure of motives, loving parent or pedophile problem mixed, I assume that D was a parent, if not the acts were questionable unsure if D was the parent, if not story sounds pretty perverted confusing-difficult to develop a clear relationship between D & P uncomfortable because you did not know the nature of the relationship a caring loving parent, or a relative, neighbor who is a molester At first I thought D might abuse P but then I realized D was just P’s parent uncomfortable with the intimacy since you aren’t aware of how they are related 149 -1 = The subject unequivocally reports negative feelings (i.e. weird, strange, uneasy, uncomfortable) or perceivedl'mterpreted the story negatively. examples: it was weird that D changed in front of P weird that D acted that way given P’s age I felt very uncomfortable D is strange D’s actions were inappropriate for D’s gender it was unusual, I didn’t like it it was disturbing, relationship was strange I was concerned something peculiar was up with D D was a bit too loving to P surprised to see D sleep with P I felt uneasy not the best environment for P D’s behavior was odd I felt a little strange uneasy D was too close to P I felt bad for P P’s parent is sick in the head. D is sexually abusing P, truly disgusting, D is white trash D is a pervert, worried for P’s safety D is molesting P disgust, urge to pummel D felt disgusted, felt bad for P unhealthy situation, D needs psychological help what D is doing could be perverted my feelings were one of child molesting D is sick I feel sick because P doesn’t know any better bad for P it sounds incestuous D was taking advantage of P strange and perverse D may be a sexually abusive parent that’s the parents role (yeah right) APPENDIX 1 Trait Rating Scales 150 APPENDIX I Aggressive Trait Rating Scale Items I_. 3.04 4.84 4.01 4.55 4.03 2.27 1.53 1.58 1.96 3.90 4.50 5.00 4.66 .97 1.09 1.89 1.97 fl. 17 l. aggressive (target trait) 16 2. ambitious 16 3. confident 70 4. independent NA 5. strong-minded (4 related and positive traits) NA 6. argumentative 2 7. belligerent NA 8. hot-headed 12 9. stubborn (4 related but negative traits) 62 10. careful 4 ll. cultured 21 12. educated 21 13. neat (4 unrelated and positive traits) 5 l4. boring 7 15. superficial l 16. superstitious 4 l7. unhealthy (4 unrelated and negative traits) L = Likableness Rating (see Anderson, 1968) Fr. = Word Frequency Rating (see Kucera & Francis, 1967) 151 (Example of aggressive trait rating scale given to students.) ForszS/AT Earlier you read a short story about a character named Donald. Please rate Donald on the following traits. If you circle 1 the trait is not at all like Donald. If you circle 10 the trait is extremely representative of Donald. Please circle only one number for each trait. ambitious not at all l....2....3....4....5....6....7....8....9....10 extremely argtanentative not at all l....2....3....4....5....6....7....8....9....10 extremely careful not at all l....2....3....4....5....6....7....8....9....10 extremely boring not at all 1....2....3....4....5....6....7....8....9....10 extremely confident not at all l....2....3....4....5....6....7....8....9....10 extremely belligerent not at all l....2....3....4....5....6....7....8....9....10 extremely cultured not at all l....2....3....4....5....6....7....8....9....10 extremely superficial not at all 1....2....3....4....5....6....7....8....9....10 extremely independent not at all l....2....3....4....5....6....7....8....9....10 extremely hot-heeded not at all l....2....3....4....5....6....7....8....9....10 extremely educated not at all l....2....3....4....5....6....7....8....9....10 extremely superstitious not at all 1....2....3....4....5....6....7....8....9....10 extremely aggressive not at all l....2....3....4....5....6....7....8....9....10 extremely strongvmiuded not at all l....2....3....4....5....6....7....8....9....10 extremely stubborn not at all l....2....3....4....5....6....7....8....9....10 extremely neat not at all 1....2....3....4....5....6....7....8....9....10 extremely tmhealthy not at all l....2....3....4....5....6....7....8....9....10 extremely 152 Nurturant Trait Rating Scale Items L NA 5.86 5.10 5.19 4.79 2.05 1.79 1.12 1.02 5.30 5.05 4.50 2.07 1.89 1.16 1.10 fl, NA 8 1 61 8 l. nurturant (target trait) 2. dependable 3. unselfish 4. friendly 5. kindly (4 related and positive traits) NA NA NA NA 6. worrier 7. oversensitive 8. bossy 9. nosey (4 related but negative traits) 1 82 16 4 10. open-minded 11. interesting 12. humorous 13. cultured (4 unrelated and positive traits) NA 1 1 3 l4. unoriginal 15. superstitious 16. uncivil 17. cowardly (4 unrelated and negative traits) L = Likableness Rating (see Anderson, 1968) Fr. = Word Frequency Rating (see Kucera & Francis, 1967) 153 (Example of nurturant trait rating scale given to students.) FonnzMS/NT Earlier you read a short story about a character named Donald. Please rate Donald on the following traits. If you circle I the trait is not at all like Donald. If you circle 10 the trait is extremely representative of Donald. Please circle only one number for each trait dependable not at all l...2....3....4....5....6....7....8....9....10 extremely worrier not at all l...2....3....4....5....6....7....8....9....10 extremely open-minded not at all l...2....3....4....5....6....7....8....9....10 extremely unoriginal not at all l...2....3....4....5....6....7....8....9....10 extremely unselfish not at all l....2....3....4....5....6....7....8....9....10 extremely oversensitive not at all l...2....3....4....5....6....7....8....9....10 extremely interesting not at all l....2....3....4....5....6....7....8....9....10 extremely superstitious not at all l...2....3....4....5....6....7....8....9....l0 exuemely friendly not at all l...2...3....4....5....6....7....8....9....10 extremely bossy not at all 1....2...3....4....5....6....7....8....9....10 extremely humorous not at all l...2....3....4....5....6....7....8....9....10 extremely uncivil not at all l....2....3....4....5....6....7....8....9....lO extremely nurturant not at all l....2....3....4....5....6....7....8....9....lO extremely kindly not at all 1....2....3....4....5....6....7....8....9....10 extremely nosey not at all l....2....3....4....5....6....7....8....9....10 extremely cultured not at all l....2....3....4....5....6....7....8....9....IO extremely cowardly not at all l....2....3....4....5....6....7....8....9....10 extremely 154 Male Heterosexualigy Trait Rating Scale Items L NA 4.43 4.01 4.10 4.56 1.02 .79 1.53 1.83 4.68 4.38 4.12 4.80 1.09 .97 1.57 1.89 EL. 59 1. sexual (target trait) 29 2. vigorous l6 3. confident NA 4. untiring ll 5. tender (4 related and positive traits) 15 6. crude 7 7. vulgar 2 8. dominating 4 9. possessive (4 related but negative traits) 27 10. productive 25 11. rational 1 12. easygoing 10 13. witty (4 unrelated and positive traits) 7 14. superficial 5 15. boring 1 l6. envious 1 17 . superstitious (4 unrelated and negative traits) L = Likableness Rating (see Anderson, 1968) Fr. = Word Frequency Rating (see Kucera & Francis, 1967) 155 (Example of male heterosexual trait rating scale given to students.) ForszS/MST Earlier you read a short story about a character named Donna Pleme rate Donna on the following traits. If you circle 1 the trait is not at all like Donna If you circle 10 the trait is extremely representative of Donna. Pleme circle only one number for each trait. vigorous not at all l....2....3....4....5....6....7....8....9....10 extremely crude not at all l....2...3....4....5....6....7....8....9....l0 extremely productive not at all l...2....3....4....5....6....7....8....9....lO exuemely superficial not at all l....2....3....4....5....6....7....8....9....10 extremely confident not at all 1....2....3....4....5....6....7....8....9....10 extremely vulgar not at all l...2....3....4....5....6....7....8....9....10 extremely rational not at all l...2....3....4....5....6....7....8....9....10 extremely boring not at all l....2....3....4....5....6....7....8....9....10 extremely untiring not at all l....2....3....4....5....6....7....8....9....10 extremely dominating not at all l....2....3....4....5....6....7....8....9....10 extremely easygoing not at all l....2...3....4....5....6....7....8....9....10 extremely envious not at all l...2....3....4....5....6....7....8....9....10 extremely sexual not at all l...2....3....4....5....6....7....8....9....10 extremely tender not at all l....2....3....4....5....6....7....8....9....10 extremely pomessive not at all 1...2....3....4....5....6....7....8....9....10 extremely wltty not at all l...2....3....4....5....6....7....8....9....10 extremely superstitious not at all l....2....3....4....5....6....7....8....9....10 extremely 156 Female Heterosexualig Trait Rating Scale Items L NA 4.40 4.48 4.56 5.22 1.59 1.27 2.24 2.23 4.50 5.00 4.32 1.89 1.18 2.13 2.24 E; 59 1. sexual (target trait) 3 2. vivacious 27 3. eager 88 4. active 67 5. warm (4 related and positive traits) 1 6. fickle 10 7. vain 5 8. inhibited ll 9. passive (4 related but negative traits) 4 10. cultured 21 11. educated 82 12. interesting 46 13. curious (4 unrelated and positive traits) l 14. superstitious 8 15. petty 5 16. impractical 2 l7. forgetful (4 unrelated and negative traits) L = Likableness Rating (see Anderson, 1968) Fr. = Word Frequency Rating (see Kucera & Francis, 1967) 157 (Example of female heterosexual trait rating scale given to students.) ForszS/FST Earlier you read a short story about a character named Donna. Pleme rate Donna on the following traits. If you circle 1 the trait is not at all like Donna. If you circle 10 the trait is extremely representative of Donna. Please circle only one number for each trait. vivacious not at all l....2....3....4....5....6....7....8....9....l0 extremely fickle not at all 1....2....3....4....5....6....7....8....9....10 extremely cultured not at all 1...2....3....4....5....6....7....8....9....10 extremely superstitious not at all 1....2....3....4....5....6....7....8....9....10 extremely eager not at all l....2....3....4....5....6....7....8....9....IO extremely vain not at all l...2....3....4....5....6....7....8....9....10 extremely educated not at all l....2....3....4....5....6....7....8....9....l0 extremely PC"! not at all 1....2....3....4....5....6....7....8....9....10 extremely active not at all l....2....3....4....5....6....7....8....9....10 extremely inhibited not at all 1....2....3....4....5....6....7....8....9....l0 extremely interesting not at all l....2....3....4....5....6....7....8....9....10 extremely impractical not at all l....2....3....4....5....6....7....8....9....10 extremely sexual not at all l....2....3....4....5....6....7....8....9....l0 extremely warm not at all l...2....3....4....5....6....7....8....9....10 extremely passive not at all l....2....3....4....5....6....7....8....9....10 extremely curious not at all 1....2....3....4....5....6....7....8....9....10 extremely forgetful not at all l....2....3....4....5....6....7....8....9....l0 extremely APPENDIX J Consent Forms 158 APPENDIX J Consent Forms Consent Form (A) Michigan State University Department of Psychology I have agreed to voluntarily participate in a research study for the collection of normative data to be used for the standardization of a neuropsychological test being developed at the Michigan State University Neuropsychological Laboratory. 1 do not knowingly suffer from any condition that causes dementia. Such a condition would contaminate my responses. All information that I provide will be kept strictly confidential and I will remain anonymous. The only information of a personal nature that I will need to provide is my age, gender and race. I will receive 1 research credit for my participation in this project. I can receive my research credit through other projects of which I have been informed. I can discontinue participation in this project at any time. Participation in this study does not guarantee any beneficial results to me. If I so desire, I can receive additional information about this project by contacting Robert Caldwell (353—4548). Signed: Date: 159 Consent Form (B) Michigan State university Department of Psychology 1. I have freely consented to participate in a scientific study being conducted by Grace Gibson, M.A. under the supervision of Anne Bogat, Ph.D. 2. I understand that I will be required to complete a short questionnaire. Participation in this study usually takes 1/2 an hour. 3. I understand that I will receive 1 research credit for my participation. 4. The study has been explained to me and I understand the explanation that has been given and what my participation will involve. 5. I am free to discontinue my participation in the study at any time. 6. I understand that I have the option of receiving research credits for my Introductory Psychology course through other projects of which I have been informed, and that my participation in the present study is voluntary. 7. I understand that the results of the study will be treated in strict confidence and that I will remain anonymous. 8. I understand that my participation in the study does not guarantee any beneficial results to me. 9. I understand that, at my request, I can receive additional explanation of the study after my participation is completed. I can contact Grace Gibson (353-4362) or Anne Bogat (353-0812) if I have any questions or concerns regarding my participation in the study. Signed: Date: APPENDIX K Feedback Sheet 160 APPENDIX K Feedback Sheet Two Short Experiments Feedback Sheet This study is an exploration of the effects of unconscious gender stereotypes, specifically aggression, sexuality, and nurturance, on student’s perceptions of a man or a woman in a sexually ambiguous situation with a child. Research on implicit stereotyping offers a theoretical explanation for understanding why the same behavior could be perceived differently when performed by men versus women. Implicit stereotyping posits that individuals have unconscious stereotypes which are easily evoked and used to judge others without the individual’s awareness. The sentences that you initially unscrambled served as a prime to activate your unconscious stereotypes about male aggression, female nurturance or male or female heterosexuality. You may also have received neutral sentences and been in the control condition. These two experiments were really part of the same project. In the second experiment you received a story with either Donald or Donna as the adult character. The sentences that you completed and the trait rating scale at the end were the dependent measures to see how the priming experiment influenced your perception of the story. Crossing out the letter ’e’ was a filler task to make the effects of the primes stronger. This experiment is investigating the effect of gender stereotypes on perceptions of female sexual abusers. Little scientific research explores the possibility of women perpetrating child sexual abuse. Researchers studying female sexual offenders have argued that sexually abusive behavior is ”masked" as normal child care behavior when committed by women, and hence, not perceived or reported. In addition to cultural taboos about female sexuality and the effects of masking, gender stereotypes could also effect the perceptions that observers have of female child molesters. Limited experimental data indicates that men and women are indeed judged differently when they sexually abuse a child; however, this is the first study to experimentally explore the effects of unconscious gender stereotypes. To evaluate the possibility that women’s sexual activity with children is masked as normal child care, your explanations for the ambiguous behavior were measured on the sentence stem form. The effects of the priming were measured on the trait rating scale. For more information about this experiment please contact: Anne Bogat, Ph.D. at 353-0812 or Grace Gibson, M.A. at 353-4362 l I LIST OF REFERENCES LI ST OF REFERENCES Abel, G. G., Gore, D. K., Holland, C. L., Camp. N., Becker, J. V., & Rathner, J. (1989). The measurement of the cognitive distortions of child molesters. Annals of Sex Research, a, 135-153. Abel, G. G., Rouleau, J. & Cunningham-Rathner, J. (1986). Sexually aggressive behavior. In W. Curran, A. McGarry & S. Shah (Eds.), Forenaig Paychiatgy aag Egyghglggy (pp. 289-313). Philadelphia: F.A. Davis Co. Allen, C. (1991). Women and man whg sexually abgaa children: A comparative analysis. Orwell, VT: The Safer Society Press. ‘ Anderson, N. H. (1968).Likab1eness ratings of 555 personality-trait words. Journal gf Persgaaligy ang Sogia; Psychology, 2(3), 272—279. Apfelberg, B., Sugar, C., & Pfeffer, A. Z. (1944). A psychiatry study of 250 sex offenders. Amaricaa Qggzaal gf Psychiatry, 100, 762-769. Araji, S., & Finkelhor, D. (1986). Abusers: A review of the research. In D. Finkelhor (Ed.), A soarcebggk on child sexual abuse (pp.89—118). Beverly Hills, CA: SAGE Pub. Banaji, M. R., & Greenwald, A. G. (1994). Implicit stereotyping and prejudice. In M. P. Zanna & J. M. Olson (Eds.), The Psychology of Prejudige: The Oggagig Sympgaigm (Vol. 7, pp 55-76). Hillsdale, NJ: Erlbaum. Banaji, M. R., & Greenwald, A. G. (1995). Implicit gender stereotyping in judgements of fame. Jgazaal gf Personality and Social Psychology, §§(2), 181-198. Banaji, M. R., Hardin, C., & Rothman, A. J. (1993). Implicit stereotyping in person judgment. Journal of Personality and Social Psychology, §§(2), 272—281. Banning, A. (1989). Mother-son incest: Confronting a prejudice. Child Abuse & Neglect, ;;, 563-570. 161 162 Bargh, H. A., Bond, R. N., Lombardi, W. J. & Tota, M. E. (1986). The additive nature of chronic and temporary sources of construct accessibility. Journal of Personality and Social Psychology, SQ(5), 869-878. Barnard, G. W., Fuller, A. K., Robbins, L., & Shaw, T. (1989). The child molester: An integrated apprgagh to evaluation and treatment. New York: Brunner/Mazel. Baron, R. S., Burgess, M. L., & Kao, C. F. (1991). Detecting and labeling prejudice: do female perpetrators go undetected? Personality and Social Psychology Bulletin, 11(2), 115-123. Bergen, D. J., & Williams, J. E. (1991). Sex stereotypes in the United States revisited: 1972-1988. Sex Roles, 23(7/8), 413-423. Blair, I. V., & Banaji, M. R. (in review). Automatic and controlled processes in gender stereotyping. Bodenhausen, G. V., & Wyer, R. S. (1985). Effects of stereotypes on decision making and information-processing strategies. Jgurnal of Personality and Social Psychology, 28(2), 267-282. Broussard, S. D., & Wagner, W. G., & Kazelskis, R. (1991). Undergraduate students’ perceptions of child sexual abuse: The impact of victim sex, perpetrator sex, respondent sex, and victim response. Joarnal of Famtly Viglaace, S(3), 267-278. Broverman, I. K., Vogel, S. R., Broverman, D. M., Clarkson, F. E., & Rosenkrantz, P. S. (1972). Sex-role stereotypes: A current appraisal. Jgurnal gf Sggial Issaea, 28(2), 59-78. Browne, A. & Finkelhor, D. (1986). Impact of child sexual abuse: A review of the research. Paychglogigal Mil—1! 221 66'77 v Brownmiller, S. (1975). Agatast gut will. New York: Simon & Schuster. Bruner, J. (1992). Another look at New Look 1. Amatigag Psychologist, 51(6). 780-783. Burgess, A. & Hartman, C. (1987). Child abuse aspects of child pornography. Psychiatric Annala, t1, 248-253. Burgess, A., Hazelwood, R., Rokous, F., Hartmen, C., & Burgess, A. (1987). Serial rapists and thair vigtims: 163 Reenactment and repetition. Presented as the New York Academy of Sciences Conference on Human Sexual Aggression: Current Perspectives, New York City. Cited in T. C. Johnson, (1989). Female child perpetrators: Children who molest other children. Child Abuse & Neglect, ;S, 571-585. Chilman, C. (1995). Quoted in S. Sleek, (1995). Safest form of sex still too shameful to discuss. American Psychological Association Monitor, February, p.8. Cole, P. M., & Woolger, C. (1989, April). The role gf emotion in the parenting difficulties of incest vigtima. Symposia presented at the meeting of the Society for Research in Child Development, Kansas City, Missouri. Condy, 8.; Templer, D.; Brown, R. & Veaco, L. (1987). Parameters of sexual contact of boys with women. Argptvaa gf Sexual Behavior, ;S, 379-394. Cooper, A. J., Swaminath, S., Baxter, D., & Poulin, C. (1990). A female sex offender with multiple paraphilias: A psychologic, physiology (laboratory sexual arousal) and endocrine case study. Canadian Jgurpal of Payghiatty, SS(4), 334-337. Cupoli, J. M., & Sewell, P. M. (1988). One thousand fifty-nine children with a chief complaint of sexual abuse. Child Abuse & Neglect, ;S, 151-162. De Francis, V. (1969). Protecting tha child vigtima pf sex crimes committed by adults. Denver, CO: The American Humane Association. DeAngelis, T. (1995). New threat associated with child abuse. The APA Monitor, ;§(4), 1, 38. Deaux, K. (1976). The behavip; gt wgmep and map. Monterey, CA: Brooks/Cole. Cited in S. T. Fiske & S. E. Taylor (1984). Social gognition, New York: Random House. Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled components. ourna f on ’ and Social Psychology, SS(1), 5-18. Dukes, R. L., & Kean, R. B. (1989). An experimental study of gender and situation in the perception and reportage of child abuse. Child Abuse & Neglect, ;S, 351- 360. Eisenberg, N., Owens, R. G., & Dewey, M. E. (1987). Attitudes of health professionals to child sexual abuse and incest. Child Abuse a Neglect, it, 109-116. 164 Ellis, K. (1990). Fatal attraction, or the post-modern prometheus. The Journal of Sex Research, g1(1), 111-121. Erdelyi, M. H. (1992). Psychodynamics and the unconscious. American Psychologist, 31(6), 784-787. Erdley, C. A. &, D’Agostino. (1988). Cognitive and affective components of automatic priming effects. Joprnal of Personality and Social Psychology, Sg(5), 741-747. Erickson W. D., Walbek, N. H., & Seely, R.K. (1987). The life histories and psychological profiles of 59 incestuous stepfathers. Bulletin of the Amerigan Academy gf Psychiatry Law, ;S(4), 349-357. Faller, K. (1987). Women who sexually abuse children. Violence and Victimology, ;, 263-276. Faller, K. (1988). The spectrum of sexual abuse in daycare: An exploratory study. Joptnal gt Eamily Violepga, S(4), 283-298. ‘ Faller, K. (1989). Characteristics of a clinical sample of sexually abused children: How boy and girl victims differ. Child Abuse & Neglect, LS, 281-291. Faller, K. (1991). Polyincestuous families: An exploratory study. Journal of Intetperagpal Vtglepga, S, 310-322. Farrell, W. (1993). T e m t of mal owe - Wh the disposable sax. New York: Simon & Schuster. Finkelhor, D. (1979). Saapally vdgttmizad childpep. New York, NY: The Free Press. Finkelhor, D. (1984). Child aaxual appaa: Naw theoty and research. New York, NY: The Free Press. Finkelhor, D. (1986). Abusers: Special topics. In D. Finkelhor (Ed.), A source book on child agapal abpaa (pp. 119—142). Beverly Hills, CA: SAGE Pub. Finkelhor, D. (1990). Early and long-term effects of child sexual abuse: An update. P 0 Research and Practica, S;, 325-330. Finkelhor, D.; Hotaling, G.; Lewis, I. & Smith, C. (1990). Sexual abuse in a national survey of adult men and women: Prevalence, characteristics, and risk factors. ghild Abuse & Neglect, ta, 19-28. 165 Finkelhor, D. & Redfield, D. (1984). How the public defines sexual abuse. In D. Finkelhor (Ed.), Child sexual abuse: New theopy and research (pp. 107-133). NY: The Free Press. Finkelhor, D. & Russell, D. (1984). Women as perpetrators: Review of the evidence. In D. Finkelhor (Ed.), Child sexual abuse: New theoty and researgh (pp. 171-187). NY: The Free Press. . Finkelhor, D.; Williams, L. with Burns, N. (1988). Nursety crimes: Sexual abuse in day cara. Newbury Park: Sage Pub. Fiske, S. T., & Taylor, S. E. (1984). Social cogp1tigp, New York: Random House. Freeman-Longo, R. (1986). The impact of sexual victimization on males. Child appaa a Neg1act, 19, 411-414. French, D. D. (1990). Distortion and lying as defense processes in the adolescent child molester. Joutpa1 pf Offender Counseling, Satyigaa & Rahapil1tatigp, 11(2), 161- 167. Freund K., & Blanchard, R. (1989). Phallometric diagnosis of pedophilia. Jourpal of Qgpaplting and glipigal Psychology, S1, 100-105. Freund, K. & Watson, R. (1991). Assessment of the sensitivity and specificity of a phallometric test: An update of phallometric diagnosis of pedophilia. Psychological Aaaesamant:A doutpal gt Cgpapltipg and Clinical Psychglggy, S, 254-260. Fritz, G. S., Stoll, K., & Wagner, N. N. (1981). A comparison of males and females who were sexually molested as children. Journal of Sax & Marital Iharapy, 1(1), 54-59. Gale, J.; Thompson, R.; Morgan, T. & Sack, W. (1988). Sexual abuse in young children: Its clinical presentation and characteristic patterns. Child appaa é Naglagt, 11, 163- 170. Garcia, L. T. (1982). Sex-role orientation and stereotypes about male-female sexuality. Saa_391aa, S(8), 863-876. Gebhard, P. H., Gagnon, J. H., Pomeroy, W. B., & Christenson, C. V. (1965). Sex offen e : An 8' types. New York: Harper & Row. 166 Gerber, G. L. (1991). Gender stereotypes and power: Perceptions of the roles in violent marriages. Sex Rolea, 21(7/8), 439-458. Giarretto, H. (1987). A comprehensive child sexual abuse treatment program. In P. Mrazek & C. H. Kempe (Eds.), Sexually abuse children and their families. Oxford: Pergamon. Gillespie, (1935). Sexual offenses against young children. Health and Empire, 1Q(Dec.), 312. Cited in L. Bender & A. Blau, (1937). The reaction of children to sexual relations with adults. American Journal pf Qrthgpayaptatpy, 1, 500-518. Goodwin, J., & DiVasto, P. (1979). Mother-daughter incest. Child Abpaa a Neglect, S, 953-957. Greenwald, A. G. (1992). New look 3: Unconscious cognition reclaimed. American Paychg1ggiat, 11(6), 766-779. Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: attitudes, self-esteem, and stereotypes. Psychological Review, 102(1), 4-27. Groth, A. N. (1978). Patterns of sexual assault against children and adolescents. In A. Burgess, A. N. Groth, L. Holstrom, & S. Sgroi (Eds.), Sexual aaaap1t pf aptldten apd adolescents. Lexington, MA: Lexington Books. Groth, A. N. with Birnhaum, J. (1979a). Map whg papa: The psychology of the affender. New York: Plenum. Groth, A. N. (1979b). Sexual trauma in the life histories of rapists and child molesters. Vigt1mglggy: Ap International Sournal, 1(1), 10-16. Groth, A. N. (1982). The incest offender. In S. Sgroi (Ed.), Handbook of alinical intervention in ahild aaapal abuse (pp.215-239). Lexington, MA: Lexington Books. Harrison-Speake, K. & Willis, F. N. (1995). Ratings of the appropriateness of touch among family members. Sgptpal of Nonverbal Behavior, 11(2), 85-100. Haugaard, J. J., & Emery, R. E. (1989). Methodological issues in child sexual abuse research. gp11d_appaa_a Neglect, 1;, 89-100. Haywood, T. W., Grossman, L. S., & Cavanaugh, J. L. (1990). Subjective versus objective measurements of deviant sexual arousal in clinical evaluations of alleged child 167 molesters. Psychological Assessment: A Journal of Consp1tipg and Clinical Psychology, 1(3), 269-275. Henderson, D. (1975). Incest. In A. Freedman, H. Kaplan & B. Sadock (Eds.), Comprehensive textbook of payghiatty, 2nd ed. Baltimore: Williams and Wilkins. Hendricks, S., Fitzpatrick, D., Hartmann, K., Quaife, M., Stratbucker, R., & Graber, B. (1988). Brain structure and function in sexual molesters of children and adolescents. Journal of C1inical Psychtatty, g2, 108-112. Herman, J. & Hirschman, L. (1981). Fathar-dapghta; incest. Cambridge, MA: Harvard University Press. Herschberger, R. (1948). Adam's pip. New York, NY:Pellegrini & Cudahy. Higgins, E. T., & King, G. (1981). Accessibility of social constructs: Information- -processing consequences of individual and contextual variability. In N. Cantor and J. F. Kihlstrom (Eds. ), P r n '0 interaction (pp. 69-121). Hillsdale, NJ: Erlbaum cited in Devine, P.G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Sgp;pal_p§_2a;appality and Social Psyahology, SS(1), 5-18. Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility and impression formation. Sgp;pa1_pfi Eaperimental Social Psychology, 11, 141-154. Hite, S. (1993). W men a r v ' n n change: The Hite Rapprts and peyond. Madison, WI: The University of Wisconsin Press. Howells, K. (1981). Adult sexual interest in children: Considerations relevant to theories of aetiology. In M. Cook and K. Howells (Eds. ), Adults eapal iptateat 1n ap11dpap, (pp. 55- 98). London: Academic Press. Hunter, J. (1991). A comparison of the psychosocial maladjustment of adult males and females sexually molested as children. Jou n nte e n ' 1e , S, 205-217. Hunter. M. (1990)._u§_e_<1_b_ox§__1th1es.lec_tesLch§1ma of sexual abuae. Lexington, MA. Lexington Books. Hyde, J. S., & Rosenberg, B. G. (1976). flal£_tpa_hpmap experience. Lexington, MA: D. C. Heath and Co. Johnson, R., & Shrier, D. (1987). Past sexual victimization by females of male patients in an adolescent 168 medicine clinic population. American Journal of Psychiatpy, 144(5), 650-652. Justice, B., & Justice, R. (1979). The broken tabog: Sex in the family. New York, NY: Human Sciences Press. Kaplan, E. A. (1990). Sex, work, and motherhood: The impossible triangle. The Journal of Sex Research, 11(3), 409-425. Kempe, R. & Kempe, C. H. (1984). The common secret: Sexual abuse of children and adolescepta. New York: W. H. Freeman and Co. Kendall-Tackett, K. & Simon, A. (1987). Perpetrators and their acts: Data from 365 adults molested as children. Child Abuse & Neglegt, 11, 237-245. Kendall-Tackett, K.; Williams, L. & Finkelhor, D. (1993). Impact of sexual abuse on children: A review and synthesis of recent empirical studies. cho i a Bulletin, 111, 164-180. Kercher, G., & McShane, M. (1984). Characterizing child sexual abuse on the basis of a multi-agency sample. Victimology, 2(3-4), 364-382. Kucera, N., & Francis, W. N. (1967). gomputatigpal anal is of resen - a r'ca ' . Providence, RI: Brown University Press. Landis, J. (1956). Experiences of 500 children with adult sexual deviation. WW ag, 91-109. Langevin, R. (1992) Biological factors contributing to paraphiliac behavior. Psychiatrig ApnaLa, 11, 307-314. Lew, M. (1988). Viatims ng longer; Men pacovetipg fpgm incest and other aexual child abuse. New York: Nevraumont Pub. Co. Lewicki, P., Hill, T., & Czyzewaka, M. (1992). Nonconscious acquisition of information. Amap1aap Psychologist, 21(6), 796-801. Lindblad, F. (1990). Child sexual abuse: The suspects, the suspicions and the background. Acta EQQQLEEEIQ Scandinavia, 12, 98-106. Margolin, L. & Craft, J. (1989). Child sexual abuse by caretakers. Family Ralations, 1S, 450-455. 169 Mathews, R., Matthews, J. K., & Speltz, K. (1989). Female sexual offenders: An exploratopy study. Orwell, VT: The Safer Society Press. Mathis, J. (1972). Clear thinking about aexual deviation. Chicago: Nelson-Hall Co. Mayer, A. (1992). Women sex offenders. Holmes Beach, FL: Learning Publications, Inc. McCarty, L. (1986). Mother-child incest: Characteristics of the offender. Child Welfare, LXV, 447- 458. Meiselman, (1978). Inces A ch 1 f causes and effects with treatment raggmmendations. San Francisco: Jossey Bass. Musk, H., & Gallagher, K. (May, 1985). Sexual and physical abuse among women inmates and their families: A national survey. Survey conducted with the support of the American Correctional Association cited in S. Travin, K. Cullen, & B. Protter (1990). Female sexual offenders: Severe victims and victimizers. Jgurnal of Fatapaig Sgiangaa, 35(1), 140—150. Naitove, C. (1988). Arts therapy with child molesters: An historical perspective on the act and an approach to treatment. The Arts in Psyahotherapy, 1S, 151-160. National Center on Child Abuse and Neglect. (1992). National Child Abuse and Neglect Data Syatem; Wgrting Eapa; 1-1990 Summaty Data Component (DHHS Publication No. ACF 92- 30361). Washington, DC: U.S. Government Printing Office. National Center on Child Abuse and Neglect. (1993). National Child Abpse apd Nagleat Qata Syatem; Working Eapa; 2-1991 Summapy Data gomponent. Washington, DC: U.S. Government Printing Office. O'Connor, A. (1987). Female sex offenders. Britiah Journal of Psychiatty, 150, 615-620. Peters, S. D., Wyatt, G. E., & Finkelhor, D. (1986). Prevalence. In D. Finkelhor (Ed.) A Sgpraabggh an Child Sexual Abuae, (pp.15-59). Beverly Hills, CA.: Sage Pub. Petrovich, M., & Templer, D. I. (1984). Heterosexual molestation of children who later became rapists. Psychological Reports, S1, 810. Plummer, K. (1981). Pedophilia: Constructing a 170 sociological baseline. In M. Cook and K. Howells (Eds.), Adult sexual interest in children (pp.221-250). New York: Academic Press Inc. Reinhart, M. (1987). Sexually abused boys. Child Sagas & Neglect, 11, 229-235. Reppucci, D. E., & Haugaard, J. J. (1989). Prevention of child sexual abuse: Myth or reality. Amerigan Psychologist, 11, 1266-1275. Risin, L. I., & Koss, M. P. (1987). The sexual abuse of boys: Prevalence and descriptive characteristics of childhood victimizations. Jgurnal gf Intatperagpal Vioiapae, 1(3), 309-323. Rothbaum, F. (1977). Developmental and gender differences in the sex stereotyping of nurturance and dominance. Devalopmental Psychglogy, 11(5), 531-532. Rowan, E.; Rowan, J. & Langelier, P. (1990). Women who molest children. Bullatin of the Ameticap Agadamy ot Psychiatty and Law, 1S, 79-83. Russell, D. (1983). The incidence and prevalence of intrafamilial and extrafamilial sexual abuse of female children. Child apuse & Neglect, 1, 133-146. Sarrel, P. M., & Masters, W. H. (1982). Sexual molestation of men by women. Archives at Saxpai Sahavigt, 11(2), 117-131. Saslawsky, D. A. & Wurtele, S. K. (1986). Educating children about sexual abuse: Implications for pediatric intervention and possible prevention. Sgp;pa1_gfi_gadiat;ia Psychology, 11, 235-245. Schacter, D. L. (1992). Understanding implicit memory: A cognitive neuroscience approach. Ameriaap Bayaholpgiat, 21(4), 559-569. Schultz, L. G., & Jones, P. (1983). Sexual abuse of children: Issues for social service and health professionals. Child Walfate, LX11(2), 99-108. Simon, L. M. J., Sales, B., Kaszniak, A., & Kahn, M. (1992). Characteristics of child molesters: Implications for the fixated-regressed dichotomy. Journal pf intetpetaopal Violence, 1(2), 211-225. Socarides, C. W. (1991). Adult-child sexual pairs: Psychoanalytic findings. The Jgurnal of Payahghiatgpy, 12, 171 185-189. Srull, T. K. &, Wyer, R. S. (1979). The role of category accessibility in the interpretation of information about persons: Some determinants and implications. Joutnai of Personality and Social Psychology, 11(10), 1660-1672. Srull, T. K. &, Wyer, R. S. (1980). Category accessibility and social perception: Some implications for the study of person memory and interpersonal judgments. Journal of Personality and Social Psychology, 1S(6), 841- 856. Tharinger, D. (1990). Impact of child sexual abuse on developing sexuality. Prpfessionai Payapolggy: Reaaarcg and Practice, 11, 331-337. Travin, S., Cullen, K., & Protter, B. (1990). Female sexual offenders: Severe victims and victimizers. Sgurpal pf Forensic Sciences, 11(1), 140-150. United States Department of Health and Human Services, National Center on Child Abuse and Neglect. (1994). Child maltreatment 1992: Re orts from s t e N ' Center on Child apusa and Neglaat. Washington, DC: U.S. Government Printing Office. Wagner, W. G., Aucoin, R., & Johnson, J. T. (1993). Psychologists' attitudes concerning child sexual abuse: The impact of sex of perpetrator, sex of victim, age of victim, and victim response. qurnal of Shild Saxuai appaa, 1(2), 61-74. Wakefield, H. & Underwager, R. (1991). Female child sexual abusers: A critical review of the literature. American Journal of Forensic Psychology, 2, 43-69. Welldon, E. V. (1988). Mother ma w r - The idealization and denigration of motharhogd. London: Free Association Books. Welldon, E. V. (1991). Psychology and psychopathology in women-a psychanalytic perspective. Bpitisp Sgprpal at Psychology, 158, 85-92. Williams, L. M., & Farrell, R. A. (1990). Legal response to child sexual abuse in day care. Ctiminai Spatige and Behavior, 11(3), 284-302. Wyatt, G. E. (1985). The sexual abuse of Afro-American and White-American women in childhood. Child apuse Q Neglect, S, 507-519. 172 Wyatt, G. E., & Peters, S. D. (1986a). Issues in the definition of child sexual abuse in prevalence research. Child Abuse & Neglect, 10, 231-240. Wyatt, G. E., & Peters, S. D. (1986b). Methodological considerations in research on the prevalence of child sexual abuse. Child Abuse & Neglect, 1Q, 241-251.