Scripts for consensual sex as risk factors for sexual aggression: A three-wave longitudinal study with university students in Germany

Abstract

Background:

Sexual aggression is a major problem among young adults. Sexual scripts are cognitive schemata representing typical elements of sexual interactions and serve as guidelines for sexual behavior. They may be linked to the risk of sexual aggression if they contain elements known to be associated with the perpetration of sexual aggression, such as alcohol use, sex with casual partners, and ambiguous communication of sexual intentions.

Objectives:

The study was designed to examine pathways from risky sexual scripts for consensual sex to risky sexual behavior and sexual aggression perpetration in men and women.

Design:

The study employed a three-wave longitudinal design with 12-month intervals. Participants were 2425 university students in Germany (58% female).

Methods:

At each wave (Time 1–Time 3), participants completed measures of risky sexual scripts and risky sexual behavior, defined by three aspects: Casual sex, alcohol consumption, and ambiguous communication of sexual intentions. Reports of sexual aggression perpetration were collected since the age of consent (14 years; Time 1) or in the past 12 months (Time 2 and Time 3) using the Sexual Aggression and Victimization Scale.

Results:

Perpetration rates for men were 9.8% at Time 1, 12.2% at Time 2, and 9.5% at Time 3. For female participants, the corresponding rates were 6.0% at Time 1, 6.3% at Time 2, and 5.1% at Time 3. The gender difference was significant at Time 1 and Time 2, but not at Time 3. As hypothesized, more risky sexual scripts prospectively predicted more risky sexual behavior, which predicted a higher risk of sexual aggression perpetration. The findings held for men and women and for participants with exclusively opposite-sex and with both opposite- and same-sex contacts.

Conclusion:

Scripts for consensual sex may be risk factors for sexual aggression among men and women if they contain elements identified as risk factors for sexual aggression. The findings suggest that changing risky sexual scripts for consensual sex may be a promising strategy for preventing sexual aggression perpetration.

Keywords

Sexual aggression sexual scripts longitudinal study university students Germany

Introduction

Sexual aggression, defined as sexual contact without consent, is a major problem among young adults in general and university students in particular. Most of the evidence on the prevalence of sexual aggression among university students comes from studies eliciting victimization reports. For example, the 2019 Campus Climate Survey including more than 180,000 students in the United States revealed that 13% of students had experienced nonconsensual contact through the use of physical force or the exploitation of their inability to consent.¹ Rates were significantly higher for women than for men, for undergraduate than for graduate students, and for sexual minorities compared with heterosexual participants. Studies from other countries also document high prevalence rates of sexual victimization in college student samples.^2,3

A smaller body of research has elicited reports of sexual aggression perpetration from men, and even fewer studies have collected perpetration reports from women.⁴ The comprehensive review of 68 North American studies by Anderson et al.⁵ found that on average, one in three male college students reported having ever engaged in behaviors defined as sexual aggression perpetration. For reports of women’s perpetration of sexual aggression against men, the review by Pina and Fisher⁶ found a range of 2% to 24%. However, as with victimization rates, prevalence rates of perpetration varied widely, not least due to methodological differences between studies.⁵

Regarding gender differences, prevalence rates for women’s perpetration are consistently lower than those of men. For example, in a study including 10 countries from the European Union, men’s rates of self-reported perpetration of sexual aggression against women ranged from 5.5% (Belgium) to 48.7% (Greece), while women’s rates of self-reported perpetration against men ranged from 2.6% (Belgium) to 14.8% (Greece).²

A comparative study of perpetration in different generation cohorts by Anderson et al.⁷ found that the gender difference in perpetration reports was nonsignificant in the youngest cohort (Millennium, born 1985–2004), whereas men had higher perpetration rates than women in the cohorts of Baby-Boomers (born 1945–1964) and Generation X (born 1965–1984). They attribute this “millennium shift” to a change in the traditional sexual script that assigns men the role of initiators and women the role of gatekeepers of sexual contact. To the extent that these roles are loosened, with women taking a more active role in initiating sexual contact and men’s use of sexual coercion becoming less tolerated, the prevalence rates for men and women might become more similar, resulting in a reduced gender gap in sexual aggression perpetration. Consistent with this reasoning, a cross-cultural study found higher rates of male victimization by an intimate partner in countries with greater gender equality.⁸ However, a series of studies from six countries (Brazil, Chile, Germany, Iran, Poland, and Turkey) in which parallel perpetration reports were collected from men and women continued to find higher perpetration rates for men than for women.⁹

Sexual scripts, sexual behavior and sexual aggression perpetration

In the search for risk factors of sexual aggression perpetration, cognitive scripts for sexual interactions have been assigned a critical role. Sexual scripts are cognitive representations of the ways in which sexual interactions are conducted, rooted in cultural constructions of gender roles and norms about sexual behavior.¹⁰ Sexual scripts are generalized across specific sexual encounters and are formed based on direct experience or observational learning, informed by prevailing cultural norms. They serve as guidelines for behavior in sexual interactions. If sexual scripts for consensual sex contain risk factors for sexual aggression, they may be relevant for the understanding of sexual aggression perpetration. In particular, three aspects of scripts for consensual sex have been linked to a higher risk of perpetration: engaging in casual sex, alcohol consumption in sexual interactions, and the ambiguous communication of sexual intentions. To the extent that these risk factors are part of an individual’s scripts for consensual sexual interactions, they may predict an increased likelihood of engaging in sexual aggression, mediated by sexual behavior congruent with the script.

As guidelines for behavior, the content of sexual scripts should be reflected in corresponding sexual behavior so that risky sexual scripts predict risky sexual behavior, with risk being defined in relation to sexual aggression perpetration. Engaging in casual sexual contacts is common among college students, as indicated by a large body of research on “hooking up.”¹¹ Casual sex may be considered a risk factor for perpetration because it creates more opportunities for engaging in sexually aggressive behavior. Impersonal sex as a risk factor of sexual aggression perpetration is part of the “confluence model” of sexual aggression supported by a substantial body of empirical research.^12,13 The longitudinal study by Zinzow and Thompson¹⁴ found that risky sexual behavior, defined by number of sexual partners, high-risk drinking, and drug use, predicted men’s perpetration of sexual aggression. Another longitudinal study with male college students found that a higher number of sexual partners and drinking in the context of sexual interactions were prospectively linked to the initiation and continuation of sexual aggression perpetration.¹⁵

However, the confluence model is designed to explain men’s use of sexual aggression, and the role of casual sex in predicting women’s sexual aggression perpetration is less well documented. This limitation is problematic especially because of the double standard applied to men’s and women’s engagement in casual sex. Whereas having many partners may enhance a man’s sexual reputation, the same behavior shown by a woman is judged as undermining her reputation.¹⁶ A large body of evidence has shown gender differences in the number of sexual partners and the frequency of hooking up in line with this double standard.^17,18 However, studies linking casual sex to sexual aggression perpetration in women are largely missing. A study with gay men and lesbian women from Germany found significant positive correlations between impersonal sex and sexual aggression perpetration in both groups.¹⁹

A second widely established behavioral risk factor for sexual aggression is alcohol use, especially in the context of sexual encounters. A review of survey results found consistent evidence for a link between alcohol use and sexual aggression perpetration, although the available studies were mostly cross-sectional.²⁰ Alcohol undermines a person’s ability to process situational information, such as the expression of nonconsent, and the awareness of personal norms prohibiting the use of aggression to obtain sexual contacts. Both general drinking habits and the habit of drinking in the context of sexual encounters were found to be associated with reports of sexual aggression perpetration. The review by Abbey et al.¹⁵ found support for assuming a causal role of alcohol from experimental studies in which participants were presented with a videotape or audiotape of a sexual assault scenario. Participants, who had received either an alcoholic or a nonalcoholic drink, then indicated how likely they would be to behave like the assailant in the scenario. Participants under the influence of alcohol were more inclined to engage in sexual aggression than sober participants.²⁰ Again, this evidence is based on findings with male participants. Alcohol use was found to be an integral part of young adults’ hooking-up scripts, indicating that risk factors for sexual aggression perpetration do not operate independently, but are part of a comprehensive mental representation of sexual encounters.²¹ A longitudinal study identified both alcohol use and having multiple sexual partners as risk factors of sexual aggression perpetration, but again, only men were included.¹⁴

Finally, the ambiguous communication of sexual intentions has been linked to an increased risk of sexual aggression perpetration. Ambiguous communication can take the form of saying “no” to a sexual advance when being willing to have sex or saying “yes” to a sexual advance when being unwilling to have sex.^22,23 In the traditional heteronormative sexual script, saying “no” despite being willing to have sex is construed as a more common behavior of women than of men in sexual encounters, and although shown across gender groups, men overestimate the use of this strategy by women.²⁴ Several studies established a link between men’s use of, and beliefs in women’s, rejection of sexual advances despite being willing to engage in sex with sexual aggression perpetration.^25,26 Like with casual sex and alcohol use, the evidence is based almost exclusively on studies with men. The ambiguous communication of sexual intentions undermines perceptions and negotiations of consent, thereby creating a context that may promote sexual advances without ascertaining consent.^27,28

Several longitudinal studies have shown that sexual scripts for consensual sexual interactions which contain these risk factors prospectively predict both sexual aggression perpetration and sexual victimization^29
–32 through their impact on sexual behavior that is congruent with the script.

This study

Studies on risk factors of sexual aggression perpetration are mostly based on cross-sectional designs, precluding the testing of directional paths from potential risk factors to sexual aggression.³³ To yield more conclusive evidence, longitudinal designs are needed in which the proposed risk factors are assessed prior to the outcomes. Moreover, evidence of sexual aggression perpetration by women is still scarce. To address these limitations, this study adopted a three-wave longitudinal design to examine the sequential paths from risky sexual scripts via risky sexual behavior to sexual aggression perpetration in both men and women. In the context of this study, risk is conceptualized specifically regarding an increased probability of engaging in sexual aggression. The study included a large sample of university students and used a validated measure of sexual aggression perpetration with behaviorally specific items, which are considered the gold standard in sexual aggression research.³⁴ Based on the theoretical arguments and empirical findings outlined above, the study tested the following predictions:

Hypothesis 1. More risky sexual scripts at Time 1 (T1), defined by the extent to which established risk factors of sexual aggression perpetration (casual sex, alcohol use, ambiguous communication) are part of a person’s cognitive script for consensual sex, predict more risky sexual behavior, defined by the same factors, at Time 2 (T2), controlling for risky sexual behavior at T1.

Hypothesis 2. More risky sexual behavior at T2 is a prospective predictor of higher odds of sexual aggression perpetration at Time 3 (T3), controlling for sexual aggression at T1 and T2.

Hypothesis 3. More risky sexual scripts at T1 are indirectly linked to a higher probability of sexual aggression perpetration at T3 through more risky sexual behavior at T2.

Although we expect women to report sexual aggression perpetration at a lower rate than men, we assume the proposed pathways from scripts to perpetration via sexual behavior to hold for both men and women. Moreover, we examined differences in relation to participants’ sexual experience background, comparing participants with exclusively opposite-sex contacts with participants with both opposite- and same-sex contacts. The group of participants with exclusively same-sex contacts was too small to be included in the comparison. Again, we expected the proposed associations to hold in a similar way across the two groups. These additional analyses were conducted to test the robustness of the associations proposed in our model.

Method

Participants

The sample consisted of 2425 university students (1010 men and 1415 women) from different state-funded, tuition-free universities in the Federal States of Berlin and Brandenburg, Germany, who were in their first year at T1 (2011). The mean age of the sample at T1 was 22.34 years (standard deviation (SD) = 2.36; range: 18–30 years). Participants were enrolled in a wide range of academic degree courses. The second and third data waves (T2, 2012, and T3, 2013) were conducted 12 and 24 months after T1. The sample size was 1685 (1033 women and 652 men) at T2 and 1619 (1000 women and 619 men) at T3. This corresponds to dropout rates of 30.5% from T1 to T2 and 3.92% from T2 to T3. Participants who dropped out after T1 had more sexual partners, were younger at first sexual intercourse, and scored higher on the risky scripts and risky behavior measures than did those who remained in the sample at T2. Thus, the data cannot be considered missing completely at random (MCAR), but can be treated as missing at random (MAR) because the information in the variables that are correlated with the presence or absence at later data waves can be used for the estimation of missing values.³⁵ Therefore, all T1 participants were included in the analyses, and missing data were handled using full information maximum likelihood (FIML) estimation, as described below.³⁶

At T1, most participants (91.7% of men and 95.1% of women) reported having had consensual sexual contacts, with 78.0% of men and 73.9% of women reported exclusively opposite-sex contacts, 9.4% of men and 20.0% of women reported sexual contacts with both opposite- and same-sex partners, and 4.3% of men and 1.2% of women reported exclusively same-sex contacts. By T3, all participants had consensual sexual experiences, with 84.4% of men and 74.3% of women reporting exclusively opposite-sex contacts, 11.8% of men and 20.0% of women reporting both opposite- and same-sex contacts, and 3.9% of men and 0.7% of women reporting exclusively same-sex contacts. The mean age at first sexual intercourse, as reported at T1, was 16.99 years for men (SD = 1.94) and 16.47 years (SD = 1.83) for women. Men had a mean number of 5.42 partners with sexual intercourse (SD = 8.02); the mean for women was 4.27 (SD = 4.00). The gender difference was significant on both variables, multivariate F(df = 22,020) = 37.78, p < .001. At T1, 86.0% of men and 91% of women reported they were currently in a steady relationship or had been in a steady relationship in the past. At T3, the corresponding rates were 93.7% for men and 95.5% for women.

Measures

Sexual aggression perpetration

Reports of sexual aggression perpetration were obtained with the Sexual Aggression and Victimization Scale (SAV-S) developed in Germany by Krahé and Berger.³⁷ The SAV-S uses behaviorally specific items pioneered in the Sexual Experiences Survey (SES) by Koss et al.^38,39 It combines (a) three coercive strategies (threat or use of physical force; exploitation of the inability of the victim to resist, e.g. due to alcohol or drug consumption; use of verbal pressure, e.g. calling the victim a failure) with (b) three victim–perpetrator relationships (current or former partner, acquaintance, and stranger) and (c) four sexual activities (sexual touch; attempted sexual intercourse; completed sexual intercourse; and other sexual acts, e.g. oral sex). This combination results in a total of 36 items. At T1, a 4-point scale of 0 (never), 1 (once), 2 (twice), and 3 (three or more times) was used for each item. Because the number of responses in the categories >2 was very low, the format was changed to a dichotomous response scale of 1 (once) and 2 (more than once) at T2 and T3, with a summary response option (I did not experience any of these actions) replacing the “0” category for each item to reduce the time needed to complete the survey. Based on filter questions about past sexual experiences (opposite-sex partners only, same-sex partners only, and both), participants received tailored versions representing different sex constellations between victims and perpetrators. For example, men who reported exclusively opposite-sex contacts received the questions about a female victim, as did women who reported exclusively same-sex contacts. At T1, participants were asked to complete the items for the time period since their 14th birthday, the legal age of consent in Germany. At T2 and T3, they were asked to complete the items for the preceding 12 months. The reliability and validity of the SAV-S were demonstrated in previous research.^40
–42

A dichotomous score was created for the analyses. Participants who endorsed none of the 36 items were assigned to the nonperpetrator category (0), and those who endorsed at least one of the items were assigned to the perpetrator category (1).¹ Participants also completed a parallel set of 36 victimization items. The findings regarding victimization are reported elsewhere.⁴³ An English translation of the SAV-S items is presented in the supplemental materials.

Risky sexual scripts

This construct was measured by a scenario-based approach developed by Krahé et al.²⁹ Participants were asked to imagine a typical situation in which they had sexual intercourse with a new partner for the first time and to rate how likely the following features would be part of that situation:¹ consumption of alcohol and degree of intoxication (six items, e.g. “How likely is it that alcohol is consumed by you; by the man/woman?”),² ambiguous communication of sexual intentions (two items addressing the tendency to say “no” when meaning “yes” by self and partner: “How likely is it that you first say ‘no’ even though you want it too?”; “How likely is it that the man/woman first says ‘no’ even though he or she wants it too”; two parallel items addressing the tendency to say “yes” when meaning “no”), and³ length of acquaintanceship and engaging in casual sex (three items, e.g. “How long have the two of you known each other?” reverse coded). A 5-point response scale was presented with options matched to the content of the items: 1 (very unlikely) to 5 (very likely) for alcohol consumption, ambiguous communication, and casual sex, 1 (not at all) to 5 (totally) for the level of intoxication, and 1 (not at all) to 5 (a few months or longer) for the length of the relationship. Responses were aggregated across the 13 items to yield a total script score. Internal consistencies were α = .69 at T1, α = .74 at T2, and α = .71 at T3.

Risky sexual behavior

Risky sexual behavior in this study is conceptualized specifically with respect to the risk of engaging in sexually aggressive behavior. Behavior reflecting risky sexual scripts was measured by seven items based on previous research.²⁹ Four items referred to alcohol use in the context of sexual interactions (e.g. “How often did you/did the other person drink alcohol in situations in which you had sexual intercourse” and “How drunk were you/was the other person in these situations?”), two items referred to ambiguous communication of sexual intentions (saying “no” when meaning “yes” and saying “yes” when meaning “no”), and one item referred to having sexual intercourse with a partner they did not know well as an indicator of casual sex (“When you had sex: how often was it with someone you knew hardly or not at all”). The 5-point response scales ranged from 1 (never/not at all drunk) to 5 (almost every time/totally drunk). At T1, participants were asked to think about situations in which they had sex “in the past.” At T2 and T3, they were instructed to think about situations in which they had sex “in the past 12 months.” Internal consistencies were α = .74 at T1, α = .72 at T2, and α = .65 at T3.

Demographics and sexual experience background

At the beginning of the questionnaire, participants were asked to indicate their gender, age, nationality, home university, and subject of study, whether they were currently in a steady relationship and whether they had been in a steady relationship in the past. In terms of sexual experience background, they were asked whether or not they had ever engaged in sexual contact with a member of the same sex and a member of the opposite sex (categories: male/female; response options: no, yes without sexual intercourse, and yes with sexual intercourse). Those who reported experience of sexual intercourse were asked to indicate their age at first intercourse and number of partners. All measures were presented in German. The full set of items for all measures may be obtained from the first author.

Procedure

Approval for the study and all materials was obtained from the Ethics Committee of the University of Potsdam, Germany (Decision No. 6/28, 4 May 2010). The study was conducted as an online survey. Invitations to participate were sent out via e-mail to first-year students of the participating higher education institutions through the respective student offices or student associations. Interested students registered in a data bank created for the purposes of this study and were sent the link to the online questionnaire upon registration. Participants were required to give informed written consent by clicking the appropriate box before being able to proceed to the items. Data were collected between January and July 2011 (T1), January and August 2012 (T2), and January and August 2013 (T3). At each data wave, participants received a 10-Euro shopping voucher for their participation. On each page of the SAV-S, participants could press a “Help” button, in case they felt the need for professional support. The help button, which triggered a message to the research team for further assistance, was pressed only once, and this turned out to be accidental.

Statistical analysis

All analyses were conducted using the Mplus software (version 8.9). The hypotheses were tested using a cross-lagged panel model (CLPM) approach. This approach examines the hypothesis that individual differences on the predictor variables can predict individual differences on the outcome variables, controlling for the stability of constructs over time and was found to yield robust findings when applied to multiple data sets examining the same associations.⁴⁴ Before testing the proposed longitudinal associations, we examined whether measurement invariance across the three data waves was given for the sexual script and sexual behavior measures, as described below. No measurement invariance was assumed for the frequency reports of sexual aggression perpetration.

To test our hypotheses, we first estimated a multigroup model in which all paths were constrained to be equal for men and women. Next, we compared this model to a model in which all paths were allowed to vary between men and women. Based on the finding that the unconstrained model did not fit the data significantly better than the constrained model, a single-group model was estimated and adopted as the final model. The single-group model included gender as a covariate to account for differences in mean scores, as described below. Missing data and non-normality of the distributions were handled by using a robust FIML estimation provided for the maximum likelihood (ML) estimator in Mplus.⁴⁵ As noted above, four T1 variables were associated with missingness: risky scripts, risky behavior, number of partners, and age at first intercourse. Risky scripts and behavior were part of the path model, and number of partners and age at first intercourse were specified as auxiliary variables, so all four variables were used for the FIML procedure. Indirect paths were tested through examining bias-corrected confidence intervals (CIs) based on 10,000 bootstraps. Access to the data on which the analyses are based can be obtained from the first author on request.

Results

Descriptive statistics, correlations, and measurement invariance

All descriptive statistics were calculated in Mplus.⁴⁶ Across all 36 items of the SAV-S, the rates of perpetration in the total sample were 7.6% at T1, 8.6% at T2, and 6.8% at T3. Broken down by gender, 9.8% of male participants at T1 endorsed at least one perpetration item for the period since their 14th birthday, 12.2% of male participants at T2 endorsed at least one item for the period of the past 12 months, and 9.5% of male participants at T3 reported at least one act of sexual aggression in the past 12 months. For female participants, the corresponding rates were 6.0% at T1, 6.3% at T2, and 5.1% at T3. The gender difference was significant at T1 and T2, but not at T3.

The means and SDs of risky sexual scripts and risky sexual behavior for the total sample and for male and female participants at each of the three data waves are presented in Table 1. Men had more risky sexual scripts than women at each wave. Women scored higher on risky sexual behavior than did men at T1, but no significant gender differences were found at T2 and T3. Broken down by sexual experience background, of participants with exclusively opposite-sex contacts, 7.4% reported sexual aggression perpetration at T1, 7.8% at T2, and 6.2% at T3. The corresponding rates for participants with both same-sex and opposite-sex contacts were 11.2% at T1, 13.8% at T2, and 9.6% at T3, which were significantly higher than the rates for exclusively heterosexual participants at each wave. The group of participants with exclusively same-sex contacts was too small to be included in the comparison.

Table 1.

Descriptive statistics for risky sexual scripts and risky sexual behavior.

Construct (range)	T1			T2			T3
Construct (range)	Total M (SD)	Men	Women	Total	Men	Women	Total	Men	Women
Risky scripts (1–5)	2.25 (0.44)	2.45* (0.40)	2.12* (0.42)	2.24 (0.44)	2.44* (0.42)	2.11* (0.41)	2.25 (0.41)	2.42* (0.39)	2.15* (0.39)
Risky behavior (1–5)	2.17 (0.56)	2.15* (0.56)	2.19* (0.55)	2.00 (0.55)	2.02 (0.57)	1.99 (0.54)	1.94 (0.55)	1.95 (0.59)	1.94 (0.51)

M: mean; SD: standard deviation.

Significant gender difference (p at least <.05) at the respective data wave. N = 2425.

The bivariate correlations between the model variables for the total sample are presented in Table 2. All correlations were significant. The stability coefficients over the three waves ranged from .64 to .69 for risky sexual scripts, from .47 to .55 for risky sexual behavior, and from .19 to .22 for sexual aggression perpetration.

Table 2.

Bivariate correlations.

		2	3	4	5	6	7	8	9
1	Risky sexual scripts T1	.42	.16	. 67	.37	.12	. 64	.36	.14
2	Risky sexual behavior T1	—	.14	.37	. 52	.08	.35	. 47	.08
3	Sexual aggression perpetration T1		—	.15	.13	. 19	.13	.14	. 22
4	Risky sexual scripts T2			—	.46	.14	. 69	.39	.14
5	Risky sexual behavior T2				—	.12	.44	. 55	.14
6	Sexual aggression perpetration T2					—	.11	.06	. 21
7	Risky sexual scripts T3						—	.48	.14
8	Risky sexual behavior T3							—	.10
9	Sexual aggression perpetration T3								—

All correlations are significant at p < .01. N = 2425. Correlations in bold and italics indicate stability coefficients within constructs.

Measurement invariance across the three waves was tested for the risky scripts and risky behavior measures. Scalar (strong) invariance could be demonstrated for the measure of sexual scripts, and metric (weak) invariance was established for the measure of risky sexual behavior. Details of these analyses are presented in the supplemental materials.

Pathways from risky sexual scripts to sexual aggression perpetration

The multigroup model in which all paths were constrained to equality between men and women fitted the data well, χ²(df = 39) = 76.89, p < .001, Comparative Fit Index [CFI] = .989, Root Mean Square Error of Approximation [RMSEA] = .028, 90% CI = [.019, .038], and Standardized Root Mean Square Residual [SRMR] = .038. Next, a model was estimated in which all paths were allowed to vary between men and women. This model also fitted the data well, χ²(df = 12) = 21.19, p = .047, CFI = .997, RMSEA = .025, 90% CI = [.003, .042], and SRMR = .011. A comparison of the CFIs for both models yielded a ΔCFI of −.008, which is below the threshold of ΔCFI of −.01 for considering one model fit superior to the other.⁴⁷ An inspection of the individual coefficients in the unconstrained model revealed that of the 27 coefficients estimated, only five differed significantly, none of which was involved in our hypotheses. Therefore, we concluded that the constrained model did not fit worse than the unconstrained model, indicating that the proposed pathways held for both men and women.

To account for gender differences in the means of risky sexual scripts, risky sexual behavior, and the percentages of sexual aggression perpetration, we next estimated a single-group model including gender as a covariate. Based on its excellent fit, χ²(df = 6) = 13.10, p = .04, CFI = .999, RMSEA = .022, 90% CI = [.004, .039], and SRMR = .008, this model was adopted as the final model, shown in Figure 1. Cross-sectional partial correlations between risky scripts and risky behavior, not shown in Figure 1 for clarity, were .47 at T1, .29 at T2, and .25 at T3; between risky scripts and sexual aggression perpetration .14 at T1, .05 at T2, and .02 at T3; and between risky behavior and sexual aggression perpetration .14 at T1, .07 at T2, and .02 at T3. The values of r ⩽ .05 were not significant.

Figure 1.

Path model for risky sexual scripts, risky sexual behavior, and sexual aggression perpetration.

First, the findings corroborate earlier research that past sexual aggression perpetration is a predictor of later perpetration. Both the direct paths from T1 to T2 and T2 to T3 and the indirect path from T1 via T2 to T3 were significant. As predicted in Hypothesis 1, risky sexual scripts at T1 prospectively predicted risky sexual behavior at T2, controlling for risky sexual behavior at T1. Consistent with Hypothesis 2, risky sexual behavior at T2 was a significant prospective predictor of sexual aggression perpetration at T3, controlling for prior perpetration at T1 and T2. Significant indirect paths (i.e. paths for which the 95% or 99% CI did not include zero) are shown in Table 3. In line with Hypothesis 3, the indirect path from risky sexual scripts at T1 to sexual aggression perpetration at T3 mediated by risky sexual behavior at T2 was significant. In addition, several other indirect paths were found. For each of the three constructs, the indirect paths from T1 to T3 via T2 were significant. Moreover, risky sexual scripts and risky sexual behavior mutually reinforced each other over time, as indicated by a significant indirect path from risky sexual scripts at T1 to risky sexual scripts at T3 via risky sexual behavior at T2 and from risky sexual behavior at T1 via risky sexual scripts at T2 to risky sexual behavior at T3.

Table 3.

Significant indirect paths: standardized coefficients.

Indirect paths	ß	CI
Risky scripts T1 → Risky behavior T2 → Perpetration T3	.014	[.003, .027]*
Risky behavior T1 → Risky behavior T2 → Perpetration T3	.034	[.007, .060]*
Perpetration T1 → Perpetration T2 → Perpetration T3	.027	[.008, .059]**
Risky scripts T1 → Risky behavior T2 → Risky scripts T3	.027	[.014, .044]**
Risky behavior T1 → Risky scripts T2 → Risky scripts T3	.059	[.035, .086]**
Risky behavior T1 → Risky behavior T2 → Risky scripts T3	.065	[.039, .093]**
Risky scripts T1 → Risky scripts T2 → Risky scripts T3	.217	[.174, .265]**
Risky behavior T1 → Risky scripts T2 → Risky behavior T3	.023	[.011, .039]**
Risky scripts T1 → Risky behavior T2 → Risky behavior T3	.068	[.040, .103]**
Risky scripts T1 → Risky scripts T2 → Risky behavior T3	.084	[.047, .127]**
Risky behavior T1 → Risky behavior T2 → Risky behavior T3	.164	[.121, .212]**

CI: confidence interval; T1: Time 1; T2: Time 2; T3: Time 3.

Bias-corrected boot-strapped confidence intervals.

95% CI; **99% CI.

To account for differences in sexual experience background, a further set of analyses compared participants with exclusively opposite-sex contacts with those reporting both opposite-sex and same-sex contacts. The multigroup model constraining the paths to be equal across the two groups showed a good fit with the data, χ²(df = 39) = 49.29, p = .13, CFI = .997, RMSEA = .015, 90% CI = [.000, .027], and SRMR = .030. It did not fit worse than the unconstrained model in which the paths were allowed to vary between the two sexual experience groups, χ²(df = 12) = 23.98, p = .02, CFI = .997, RMSEA = .030, 90% CI = [.011, .048], SRMR = .012, and ΔCFI = .00. Therefore, despite differences in prevalence rates, there was no indication that the proposed paths would be different in the two groups differing in sexual experience background.

Discussion

This study adopted a three-wave longitudinal design to examine risky sexual scripts for consensual sex as a prospective predictor of sexual aggression perpetration, mediated by risky sexual behavior. With this design, we were able to conduct a proper test of the proposed mediational model, which requires risky sexual scripts to be assessed prior to risky sexual behavior, and risky sexual behavior prior to sexual aggression perpetration, controlling for the stability of the predictor, mediator, and outcome variables over time. The study included a large sample of male and female university students and used a validated measure of sexual aggression perpetration with behaviorally specific items.

In evaluating the findings of this study, it is important to bear in mind that the data were collected from 2010 to 2013. In the years since, the understanding of sexual violence and research designs have become broader, both in terms of questioning the cisgender focus of past sexual aggression research and increasing the study of sexual aggression in sexual minorities and in terms of raising awareness about sexual aggression through the #Metoo movement.^48
–50 These changes may affect the interpretation of the prevalence rates of sexual aggression perpetration found in our study, as discussed below. They are less critical for the interpretation of the associations of sexual scripts and sexual behavior with the odds of sexual aggression perpetration. Recent research on sexual scripts has shown that the traditional heterosexual script remains widely endorsed and that the association between sexual scripts and sexual aggression perpetration held across different sexual orientations.^51,52 Moreover, recent research on sexual behavior, especially with college students, has shown that the aspects defining risky sexual behavior for the purposes of this study, such as alcohol use, casual sex, and ambiguous communication in sexual interactions, remain prominent in young adults’ sexual interactions.^21,53 Finally, there is little evidence of a decrease in rape myth acceptance and victim-blaming after the start of the #MeToo movement.^54,55 Therefore, we believe our findings may contribute to a better understanding of risk factors of men’s and women’s perpetration of sexual aggression in the current research landscape and societal debate.

Perpetration rates for male participants were in the region of 10% in the three waves of this study. This rate is lower than the mean rate of 25% found in the meta-analysis of studies from the United States and Canada conducted between 2000 and 2017 by Anderson et al.⁵ that used the SES.³⁹ Because the SES is comparable in breadth to the SAV-S used in this study, it is unlikely that differences in methodology are the main reason for the lower rates. Later studies from Germany using the SAV-S have yielded higher rates of 17%³ and 19%.⁵⁶ For female perpetration, no comparable meta-analytic findings are available, but the review of studies by Anderson et al.⁷ shows a substantial range in which the current rates of 5% to 6% would fall in the lower range. It is also worth noting that the prevalence rates were higher for men, and not lower for women, in the 12-month periods covered at T2 and T3 than for the longer period since their 14th birthday covered at T1. This finding is consistent with other research showing that being at university, especially the first year, is a risky time period with regard to sexual aggression.⁵⁷

Our study also showed that perpetration reports were significantly correlated across the three waves and that the indirect association of perpetration at T1 with perpetration at T3 was also significant. This pattern is consistent with a meta-analysis of longitudinal studies in which past perpetration emerged as the strongest predictor of men’s sexually aggressive behavior.⁵⁸

The hypothesized pathways from risky sexual scripts via risky sexual behavior to sexual aggression perpetration, controlling for past sexual aggression, were fully confirmed by the data. Although we found gender differences in perpetration rates at T1 and T2, in mean risky sexual script scores at all three waves, and in risky sexual behavior at T1, the associations between risky sexual scripts, risky sexual behavior, and sexual aggression perpetration were not moderated by gender. Furthermore, no evidence of a moderation effect of sexual experience background was found, with the model fitting participants with exclusively opposite-sex contacts in the same way as participants with both opposite- and same-sex contacts.

The findings support the role of the three categories of risk factors examined as part of participants’ scripts for consensual sexual interactions. A recent meta-analysis identified frequent hook-ups and alcohol use as significant risk markers of male college students’ perpetration of sexual aggression.⁵⁹ A study by Walsh et al.⁶⁰ established significant associations between men’s use of ambiguous consent communication and sexual aggression perpetration. By including women, this study corroborates the importance of these variables for understanding risk factors of the perpetration of sexual aggression across gender groups. The correlations between sexual scripts and sexual behavior are substantial, not only cross-sectionally but also over time, supporting the conceptualization of scripts as guidelines for behavior.

Strengths and limitations

Our study went beyond past research in several respects: by including a large sample of university students from Germany, studying men’s and women’s perpetration, considering participants’ sexual experience background in terms of exclusively opposite-sex and both opposite- and same-sex contacts, using a three-wave longitudinal design with 12-month intervals, and applying the theoretical concept of cognitive scripts for consensual sex as well as established risk factors of perpetration to the prediction of sexual aggression perpetration.

At the same time, some limitations must be noted. First, although the sample was large, it was a convenience sample of university students and not based on random sampling. Second, the number of participants with exclusively same-sex contacts was too small to be included in the comparative analysis. Third, although our measure referred to individuals with both same-sex and opposite-sex sexual contacts, the differentiation between men and women did not give room for the inclusion of individuals with a nonbinary gender identification. Fourth, although the retention rate at T3 of 67% of the original sample can be considered satisfactory, dropout status was associated with some of the variables of interest. To deal with this problem and retain all T1 participants in the sample, appropriate procedures for dealing with missing data were employed. Finally, other established risk factors of sexual aggression perpetration were not addressed in the study, such as rape myth acceptance,⁶¹ hostile attitudes toward women,^62,63 or peer acceptance of sexual aggression.⁶⁴

Conclusion

Our findings suggest that changing risky sexual scripts for consensual sex may be a promising strategy for preventing sexual aggression perpetration. These interventions should address the risk factors for nonconsensual sex identified in this study, such as promoting the understanding and negotiation of sexual consent,⁶⁵ as well as educating about the risk of casual sex and alcohol use. A recent longitudinal study including a baseline assessment, postintervention, and two follow-up measures covering a total of 23 months showed that risky sexual scripts may be reduced significantly through a theory-based intervention, leading to a reduction in the risk of sexual aggression perpetration.^66,67 Recent years have seen a rapid growth of programs for promoting bystander interventions.⁶⁸ Changing patterns of sexuality-related cognitions and behavior is an important task for preempting situations in which bystanders might have to intervene to stop potential perpetrators and protect potential victims.

Supplemental Material

sj-docx-1-whe-10.1177_17455057231213269 – Supplemental material for Scripts for consensual sex as risk factors for sexual aggression: A three-wave longitudinal study with university students in Germany

Supplemental material, sj-docx-1-whe-10.1177_17455057231213269 for Scripts for consensual sex as risk factors for sexual aggression: A three-wave longitudinal study with university students in Germany by Barbara Krahé and Anja Berger in Women’s Health

Footnotes

Declarations

ORCID iD

Barbara Krahé

Supplemental material

Supplemental material for this article is available online.

References

Cantor

Fisher

Chibnall

, et al. Report on the AAU campus climate survey on sexual assault and sexual misconduct (revised), 2020, https://www.aau.edu/sites/default/files/AAU-Files/Key-Issues/Campus-Safety/Revised%20Aggregate%20report%20%20and%20appendices%201-7_(01-16-2020_FINAL).pdf

Krahé

Berger

Vanwesenbeeck

, et al. Prevalence and correlates of young people’s sexual aggression perpetration and victimisation in 10 European countries: a multi-level analysis. Cult Health Sex 2015; 17(6): 682–699.

Krahé

Schuster

Tomaszewska

Prevalence of sexual aggression victimization and perpetration in a German university student sample. Arch Sex Behav 2021; 50(5): 2109–2121.

Hales

ST.

Sexual violence in higher education. In: Humphreys

Towl

(eds) Stopping gender-based violence in higher education. London: Routledge, 2022, pp. 109–127.

Anderson

Silver

Ciampaglia

, et al. The frequency of sexual perpetration in college men: a systematic review of reported prevalence rates from 2000 to 2017. Trauma Violence Abus 2021; 22(3): 481–495.

Fisher

Pina

An overview of the literature on female-perpetrated adult male sexual victimization. Aggress Violent Beh 2013; 18(1): 54–61.

Anderson

Struckman-Johnson

Smeaton

GL.

Generation by gender differences in use of sexual aggression: a replication of the millennial shift. J Sex Res 2021; 58(3): 383–395.

Hines

DA.

Predictors of sexual coercion against women and men: a multilevel, multinational study of university students. Arch Sex Behav 2007; 36(3): 403–422.

Krahé

Schuster

Tomaszewska

Men’s sexual victimization by women: a neglected problem. In: Bartsch

Krieg

Schuchmann

, et al. (eds) Gender & crime. Baden-Baden: Nomos Verlagsgesellschaft, 2022.

10.

Simon

Gagnon

JH.

Sexual scripts: permanence and change. Arch Sex Behav 1986; 15(2): 97–120.

11.

Roberson

Olmstead

Fincham

FD.

Hooking up during the college years: is there a pattern. Cult Health Sex 2015; 17(5): 576–591.

12.

Malamuth

Hald

. The confluence mediational model of sexual aggression. In: Craig

Rettenberger

Beech

, et al. (eds) The Wiley handbook on the theories, assessment and treatment of sexual offending. Chichester: Wiley Blackwell, 2017, pp. 53–71.

13.

Yucel

Angelone

Jones

MC.

Reassessing the confluence model of men’s risk for sexual aggression. J Interpers Violence 2023; 38(7–8): 6062–6084.

14.

Zinzow

Thompson

A longitudinal study of risk factors for repeated sexual coercion and assault in US

Arch Sex Behav 2015; 44(1): 213–222.

15.

Abbey

Wegner

Pierce

, et al. Patterns of sexual aggression in a community sample of young men: risk factors associated with persistence, desistance, and initiation over a one year interval. Psychol Violence 2012; 2(1): 1–15.

16.

Endendijk

van Baar

Deković

He is a stud, she is a slut! A meta-analysis on the continued existence of sexual double standards. Pers Soc Psychol Rev 2020; 24(2): 163–190.

17.

Petersen

Hyde

JS.

A meta-analytic review of research on gender differences in sexuality, 1993–2007. Psychol Bull 2010; 136(1): 21–38.

18.

Napper

Montes

Kenney

, et al. Assessing the personal negative impacts of hooking up experienced by college students: gender differences and mental health. J Sex Res 2016; 53(7): 766–775.

19.

Waldis

Herzberg

Herold

, et al. Predictors of sexual aggression among gay men and lesbian women: an application of Malamuth’s confluence model. Aggress Behav 2023; 49(2): 154–164.

20.

Abbey

Wegner

Woerner

, et al. Review of survey and experimental research that examines the relationship between alcohol consumption and men’s sexual aggression perpetration. Trauma Violence Abus 2014; 15(4): 265–282.

21.

Garcia

Litt

Davis

, et al. Growing up, hooking up, and drinking: a review of uncommitted sexual behavior and its association with alcohol use and related consequences among adolescents and young adults in the United States. Front Psychol 2019; 10: 1872.

22.

Canan

Jozkowski

Crawford

BL.

Sexual assault supportive attitudes: rape myth acceptance and token resistance in Greek and Non-Greek college students from two university samples in the United States. J Interpers Violence 2018; 33(22): 3502–3530.

23.

Muehlenhard

Hollabaugh

LC.

Do women sometimes say no when they mean yes? The prevalence and correlates of women’s token resistance to sex. J Pers Soc Psychol 1988; 54(5): 872–879.

24.

Emmers-Sommer

TM.

Do men and women differ in their perceptions of women’s and men’s saying “no” when they mean “yes” to sex?: an examination between and within gender. Sex Cult 2016; 20(2): 373–385.

25.

Krahé

Scheinberger-Olwig

Kolpin

Ambiguous communication of sexual intentions as a risk marker of sexual aggression. Sex Roles 2000; 42(5/6): 313–337.

26.

Loh

Gidycz

Lobo

, et al. A prospective analysis of sexual assault perpetration: risk factors related to perpetrator characteristics. J Interpers Violence 2005; 20(10): 1325–1348.

27.

Schipani-McLaughlin

Leone

Salazar

, et al. Knowledge of sexual consent as a protective factor against sexual violence perpetration among first-year college men: a moderation analysis. Health Educ Res 2023; 38(2): 139–149.

28.

Moyano

Sánchez-Fuentes

MDM

Parra-Barrera

, et al. Only “yes” means “yes”: negotiation of sex and its link with sexual violence. J Interpers Violence 2023; 38(3–4): 2759–2777.

29.

Krahé

Bieneck

Scheinberger-Olwig

The role of sexual scripts in sexual aggression and victimization. Arch Sex Behav 2007; 36(5): 687–701.

30.

D’Abreu

LCF

Krahé

. Predicting sexual aggression in male college students in Brazil. Psychol Men Masc2014; 15(2): 152–162.

31.

Schuster

Krahé

Predictors of sexual aggression perpetration among male and female college students: cross-cultural evidence from Chile and Turkey. Sex Abus 2019; 31(3): 318–343.

32.

Tomaszewska

Krahé

Predictors of sexual aggression victimization and perpetration among Polish university students: a longitudinal study. Arch Sex Behav 2018; 47(2): 493–505.

33.

O’Connor

McMahon

Cusano

, et al. Predictors of campus sexual violence perpetration: a systematic review of research, sampling, and study design. Aggression Violent Behavior 2021(58): 101607.

34.

Cook

Gidycz

Koss

, et al. Emerging issues in the measurement of rape victimization. Violence Against Women 2011; 17(2): 201–218.

35.

Nicholson

Deboeck

Howard

Attrition in developmental psychology. Int J Behav Dev 2017; 41(1): 143–153.

36.

Enders

CK.

Applied missing data analysis. New York: Guilford Press, 2010.

37.

Krahé

Berger

Men and women as perpetrators and victims of sexual aggression in heterosexual and same-sex encounters: a study of first-year college students in Germany. Aggress Behav 2013; 39(5): 391–404.

38.

Koss

Gidycz

Wisniewski

The scope of rape: incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. J Consult Clin Psychol 1987; 55(2): 162–170.

39.

Koss

Abbey

Campbell

, et al. Revising the SES: a collaborative process to improve assessment of sexual aggression and victimization. Psychol Women Quart 2007; 31(4): 357–370.

40.

Krahé

Haas

de Vanwesenbeeck

, et al. Interpreting survey questions about sexual aggression in cross-cultural research: a qualitative study with young adults from nine European countries. Sex Cult 2016; 20(1): 1–23.

41.

Tomaszewska

Schuster

Marchewka

, et al. Order effects of presenting coercive tactics on young adults’ reports of sexual victimization. J Interpers Violence 2022; 37(19-20): NP17784–NP17806.

42.

Marchewka

Tomaszewska

Schuster

, et al. Unacknowledged and missed cases of sexual victimization: a comparison of responses to broad versus behaviorally specific questions. Aggress Behav 2022; 48(6): 573–582.

43.

Krahé

Berger

Pathways from college students’ cognitive scripts for consensual sex to sexual victimization: a three-wave longitudinal study. J Sex Res 2021; 58(9): 1130–1139.

44.

Orth

Clark

Donnellan

, et al. Testing prospective effects in longitudinal research: comparing seven competing cross-lagged models. J Pers Soc Psychol 2021; 120(4): 1013–1034.

45.

Muthén

Asparov

Estimator choices with categorical outcomes, 2015, https://www.statmodel.com/download/EstimatorChoices.pdf

46.

Muthén

BO.

Mplus user’s guide (8th ed.), 1998–2017, https://www.statmodel.com/ugexcerpts.shtml

47.

Cheung

Rensvold

RB.

Evaluating goodness-of-fit indexes for testing measurement invariance. Struct Equ Modeling 2002; 9(2): 233–255.

48.

Fileborn

Loney-Howes

Introduction: mapping the emergence of #MeToo. In: Fileborn

Loney-Howes

(eds) #MeToo and the politics of social change. Cham: Springer International Publishing, 2019, pp. 1–18.

49.

Jaffe

Cero

DiLillo

The #MeToo movement and perceptions of sexual assault: college students’ recognition of sexual assault experiences over time. Psychol Violence 2021; 11(2): 209–218.

50.

Trottier

Nolet

Benbouriche

, et al. Sexual violence perpetration and victimization: providing prevalence rates for understudied populations. Violence Gender 2021; 8(2): 59–66.

51.

Fournier

Pathak

Hoffmann

, et al. A comparison of sexual minority and heterosexual college students on gendered sexual scripts and sexual coercion perpetration. J Interpers Violence 2023; 38(7–8): 6167–6194.

52.

Papp

Ward

Marshall

RA.

Contributions of reality TV consumption to college women’s endorsement of the heterosexual script and acceptance of sexualized aggression. Psychol Women Quart 2021; 46: 50–65.

53.

Tomaszewska

Schuster

Krahé

Young adults’ understanding of sexual competence: a qualitative study with German university students. Sex Res Soc Policy 2023; 20(1): 145–159.

54.

Leary

MG.

Is the #MeToo movement for real? Implications for jurors’ biases in sexual assault cases. La Law Rev 2020; 81(1): 82–118.

55.

Maier

SL.

Rape victim advocates’ perceptions of the #MeToo movement: opportunities, challenges, and sustainability. J Interpers Violence 2023; 38(1–2): NP336–NP365.

56.

Schuster

Tomaszewska

Marchewka

, et al. Does question format matter in assessing the prevalence of sexual aggression? a methodological study. J Sex Res 2021; 58(4): 502–511.

57.

Campbell

Sabri

Budhathoki

, et al. Unwanted sexual acts among university students: correlates of victimization and perpetration. J Interpers Violence 2021; 36(1–2): NP504–NP526.

58.

Steele

Martin

Yakubovich

, et al. Risk and protective factors for men’s sexual violence against women at higher education institutions: a systematic and meta-analytic review of the longitudinal evidence. Trauma Violence Abuse 2022; 23(3): 716–732.

59.

Spencer

Rivas-Koehl

Astle

, et al. Risk markers for male perpetration of sexual assault on college campuses: a meta-analysis. Trauma Violence Abuse 2023; 24(4): 2486–2497.

60.

Walsh

Sarvet

Wall

, et al. Prevalence and correlates of sexual assault perpetration and ambiguous consent in a representative sample of college students. J Interpers Violence 2021; 36(13–14): NP7005–NP7026.

61.

O’Connor

McMahon

The role of rape myths in contributing to sexual aggression: theory, research, and implications for prevention. In: Orchowski

Berkowitz

(eds) Engaging boys and men in sexual assault prevention. Amsterdam: Elsevier, 2022, pp. 97–123.

62.

Hales

Gannon

TA.

Understanding sexual aggression in UK male university students: an empirical assessment of prevalence and psychological risk factors. Sex Abuse 2022; 34(6): 744–770.

63.

Ray

Parkhill

MR.

Components of hostile masculinity and their associations with male-perpetrated sexual aggression toward women: a systematic review. Trauma Violence Abuse 2023; 24(2): 355–368.

64.

Dean

Swartout

KM.

Sexual assault perpetration. In: Geffner

White

Hamberger

(eds) Handbook of interpersonal violence and abuse across the lifespan. Cham: Springer International Publishing 2020, pp. 1–24.

65.

Ortiz

Shafer

Unblurring the lines of sexual consent with a college student-driven sexual consent education campaign. J Am Coll Health 2018; 66(6): 450–456.

66.

Schuster

Tomaszewska

Krahé

A theory-based intervention to reduce risk and vulnerability factors of sexual aggression perpetration and victimization in German university students. J Sex Res 2023; 60(8): 1206–1221.

67.

Tomaszewska

Schuster

Krahé

Evaluating a theory-based online program for preventing sexual aggression: an experimental-longitudinal study with German university students. Sex Abuse 2023; 35(8): 953–980.

68.

Park

Kim

S-H.

A systematic review and meta-analysis of bystander intervention programs for intimate partner violence and sexual assault. Psychol Violence 2023; 13(2): 93–106.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB