Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Oct 18.
Published in final edited form as: Psychol Violence. 2024 Oct 3;15(2):235–246. doi: 10.1037/vio0000550

An Experimental Comparison of Three Phrases Used to Convey Lack of Consent: How Does Language Affect Men’s Self-Reports of Sexual Aggression Against Women?

Antonia Abbey 1, M Colleen McDaniel 1, Elise VanParis 1, Breanne R Helmers 1
PMCID: PMC12520174  NIHMSID: NIHMS2056305  PMID: 41098620

Abstract

Objective:

This experiment expands the findings from Abbey et al. (2021) by randomly assigning participants to one of three versions of a sexual aggression measure that differed only in the language used to convey lack of consent: (1) “make her”; (2) “without her consent”; or (3) “when she didn’t want to.”

Method:

Men between the ages of 18 and 35 (N = 1291) were recruited through a Qualtrics Panel for a study of their dating and sexual experiences with women. They completed demographics; validity indicators; and one randomly assigned version of the sexual aggression measure.

Results:

Self-reported rates of sexual aggression were high across conditions. Participants who responded to questions that used “make her” language reported significantly higher rates of verbally coerced penetrative sex than did participants who responded to questions about sex “without her consent” or “when she didn’t want to.” In contrast, self-reported rates of completed rape (i.e., physical force or incapacitation tactics), attempted rape, and nonpenetrative sexual contact did not significantly differ based on the language used to convey lack of consent. The magnitude of the correlations between the total number of sexually aggressive acts and convergent and discriminant validity measures was comparable regardless of the phrasing used to convey lack of consent.

Conclusions:

Tacit societal approval of the use of verbal pressure to obtain sex makes it difficult for people to recognize their own use of these tactics. Societal-level solutions are required to create a climate in which freely given consent is the norm.

Keywords: sexual aggression, perpetration, sexual assault, rape, measurement, consent, experiment, Sexual Experiences Survey

An Experimental Comparison of Three Phrases Used to Convey Lack of Consent: How Does Language Affect Men's Self-Reports of Sexual Aggression Against Women?

Sexual assaults are rarely reported to the authorities (Reich et al., 2022; Spohn, 2020; Wolitzky-Taylor et al., 2011). Consequently, self-report surveys are crucial for obtaining information about perpetrators, including assessing prevalence, identifying potentially malleable causes of their behavior, and developing effective, evidence-based primary prevention programs. To our knowledge, Kanin (1967) provided the first self-report data on perpetration, finding that 25.5% of a random sample of unmarried men at a large university had engaged in sexual aggression while on a date with a woman. The most commonly used measure of sexual assault is the Sexual Experiences Survey (SES), which was first published in the early 1980s and has gone through several revisions (Koss et al., 2007; Koss et al., 1987; Koss & Oros, 1982). In the parallel perpetration and victimization versions of this measure, the SES uses behaviorally-specific language to describe sexual activities and the tactics used to obtain them.1 The SES avoids using terms such as “sexual assault” or “rape” in the items because they are often not clearly understood, are likely to elicit socially desirable responses, and because legal definitions vary across jurisdictions and time (Cook et al., 2011; Koss et al., 2007; Spohn, 2020). This paper follows common conventions in the field and uses the terms “sexual assault” to refer to all types of nonconsensual sexual contact. Items in the SES and other measures of sexual assault have often been divided into the subcategories of nonpenetrative sexual contact, verbally coerced penetrative sex, attempted rape, and completed rape (Koss et al., 2007; Koss et al., 1987). Attempted and completed rape include the subset of acts that involve penetrative sex obtained (or attempted) through the use of physical force, threats of physical force, or with someone incapacitated and unable to consent (see the next section for definitions of consent). Verbal coercion involves the use of tactics such as persistent pressure, displays of negative affect, verbal threats, and/or lies to obtain sex without freely given consent. These acts are not illegal in most jurisdictions; nonetheless, they cause harm to victims by violating their right to control their bodily integrity and sexual autonomy (United Nations Human Rights Council, 2021). Nonpenetrative sexual contact involves sexual touching or kissing without consent and can be a sex crime in some jurisdictions depending on the tactics used (Spohn, 2020). In this paper, the term “sexual aggression” refers to perpetration using the range of sex acts and tactics described in this paragraph.

Despite the challenges associated with developing a self-report measure of sexual aggression that is well understood by people from a wide range of backgrounds, the perpetration version of the SES has demonstrated good convergent and discriminant validity (Abbey et al., 2021; Anderson et al., 2017; Johnson et al., 2017; Ouimette et al., 2000; Testa et al., 2004). Some concerns have also been identified that suggest areas for further refinement. For example, several research teams have compared participants’ responses to the perpetration version of the SES with their responses to other behaviorally-specific sexual aggression measures and found less overlap than anticipated (Cook, 2002; Strang & Peterson, 2017; Strang et al., 2013). Many myths and misperceptions of sexual assault persist (Wilson & Miller, 2016); thus it is not surprising that minor changes in language can impact participants’ recall, interpretation, and willingness to disclose that they committed these acts.

The experiment described in this manuscript builds on past research by examining how the language used to convey lack of consent influences self-reported sexual aggression. This study focused specifically on men’s sexual aggression of women because men’s violence against women occurs at extremely high rates and has a long history that is embedded in law, cultural traditions, and many religions (De Coster & Heimer, 2021; Hunnicutt, 2009; World Health Organization, 2013). The following sections of the paper provide a brief overview of how consent has been assessed in past research, findings from relevant past methodological research, and further description of this study’s aims.

Definitions of Consent and Assessment of Consent in Sexual Assault Surveys

Consent is defined as giving assent or approval (Merriam-Webster Dictionary, n.d.). Some researchers have observed that there is a gap between people’s definitions of sexual consent and the strategies they report using to assess it in their own sexual interactions (Jozkowski et al. 2014; Muehlenhard et al., 2016). When asked how they assess their own partner’s consent, young adults report that they rely primarily on nonverbal cues (Salazar et al., 2017; Schobert et al., 2021; Shumlich & Fisher, 2020). Researchers have also considered the distinction between wanting or desiring some sexual activity and consenting to it, noting that someone may feel sexual desire but still not consent to sex, perhaps because it is against their religion or because they do not feel comfortable with their partner (Muehlenhard et al., 2016). Another important distinction is between not wanting to engage in some sexual activity but freely consenting to it, perhaps to show love to one’s partner. Furthermore, not wanting sex but freely consenting needs to be distinguished from giving up resistance but not consenting (i.e., compliance; capitulation). Verbal coercion is often intended to wear someone down, so that they reluctantly consent or quit resisting despite their lack of consent. The term “freely given consent” has been increasingly used in the literature to convey that when someone complies due to pressure or threats, they have not truly consented (Adams-Curtis & Forbes, 2004; Pugh & Becker, 2018).

The 1987 version of the SES (Koss et al., 1987) was the gold standard for measuring self-reported victimization and perpetration for 20 years until the publication of the revised SES (Koss et al., 2007). Two phrases were used to convey lack of consent in the 1987 perpetration version, “when the woman didn’t want to” and “to make her.” Some items used both of these phrases and some used only one. The “when she didn’t want to” language was adopted by Abbey and colleagues when they developed a modified SES with added questions about alcohol and drug tactics to address specific research questions (Abbey et al., 1998; 2006; 2011; 2018). The “make her” language was adopted by Abbey, Parkhill, and Koss (2005) when they developed a modified SES that examined how organizing questions by tactics versus by sexual activities impacted self-reported rates of victimization and perpetration. Krahé and Berger’s (2013) Sexual Aggression Scale also built on the SES and uses the phrase “made another person” to convey lack of consent.

Koss et al. (2007) changed the language used to convey lack of consent in the revised SES to, “without (my/their) consent.” Koss et al. did not provide empirical evidence for this modification; instead they provided the rationale that “‘when you didn’t want to” does not establish that nonconsent was expressed” (p. 359). At that time, the laws in most jurisdictions used a consent standard that requires victims to say or do something to indicate their nonconsent; putting the onus of responsibility on victims to demonstrate their lack of consent rather than on perpetrators to obtain consent (as most laws still do). Koss et al. made this change to the SES consent language to silence critics who argued that the SES exaggerated sexual assault prevalence by including incidents in which victims did not make their lack of consent clear to the perpetrator (Rutherford, 2017). Abbey and colleagues (Abbey et al., 2006; 2011) contend that this concern does not apply to the perpetration version of the SES. When someone agrees that they did something sexual to another person when that other person didn’t want to, endorsement of the item demonstrates that the instigator knew that the other person did not freely consent to that sexual activity. Laws do not punish all aggressive or unethical behavior; however, most researchers and advocates would argue that such behavior is an act of sexual aggression (Rutherford, 2017; Spohn & Horney, 1992; United Nations Human Rights Council, 2021).

Experimental Research Comparing Different Sexual Assault Perpetration Measures

In an online survey with a TurkPrime sample, Abbey et al. (2021) randomly assigned 938 male participants to complete one of three versions of the SES that had been used in published research to assess sexual assault perpetration: (1) Koss et al. (2007); (2) Abbey, Jacques-Tiura, and LeBreton (2011); and (3) Abbey, Parkhill, and Koss, (2005). These three measures share a common core of tactics and sexual activities; however they differ on multiple dimensions including the total number of items, the number of items that assess each type of tactic, how the items are formatted, and the language used to convey lack of consent (described in the next section). The authors compared the overall rates of perpetration since age 14, as well as the rates for the most severe type of sexual aggression committed: none, nonpenetrative sexual contact, verbally coerced penetrative sex, attempted rape, or completed rape. There were significant differences in overall rates of sexual aggression, which were primarily due to differences in rates of verbally coerced penetrative sex. Participants who completed the Abbey et al. (2005) measure reported significantly higher rates of verbally coerced penetrative sex (21.4%) than did participants who completed the Koss et al. (2007) measure (5.7%) or the Abbey et al. (2011) measure (13.3%). No significant differences were found between the three measures in the percentage of men who reported committing an act of completed rape (range: 9.7% - 14.1%). All three versions of the SES had similar patterns of correlations with convergent (e.g., hostility toward women, friends’ approval of forced sex, heavy episodic alcohol consumption) and discriminant (e.g., life satisfaction, perceived stress) validity indicators.

Exploratory analyses suggested that these differences might be explained by the fact that the various verbally coercive tactics were separated into four items in the Abbey et al. (2005) measure, as compared to being combined into two items in the Koss et al. (2007) and Abbey et al. (2011) measures. Combining a long list of strategies in a single item may lead to underreporting, either because participants skimmed the list and missed the strategy that corresponded to their past behavior; or because they believed all of the strategies had to apply for them to endorse the item; or because some of the strategies sounded worse than what they did so they did not want to endorse the item (Strang & Peterson, 2017). Although this post hoc explanation is plausible, there were other differences between the three measures, including the consent language, that might also explain the significant differences in rates of verbally coerced penetrative sex.

This Study’s Goals and Research Questions

The experiment described in this paper builds on and expands the findings from Abbey et al. (2021) by randomly assigning participants to one of three versions of a measure of sexual aggression that differed only in the language used to convey lack of consent. Abbey et al. (2005) served as the foundation (in terms of the tactics first format and the number of items included to assess each type of tactic) because it produced the highest overall sexual aggression prevalence rate in Abbey et al. (2021). Thus, participants answered the exact same questions except the language embedded within them to convey lack of consent was either (1) the “make her” language from Abbey et al. (2005); (2) the “without consent” language from Koss et al. (2007); or (3) the “when she didn’t want to” language from Abbey et al. (2011).2

The primary goal of this study was to evaluate the validity of these three measures in a between-person randomized experimental design. Randomization ensures that participants systematically differ only on the independent variable, which in this study was the language used to convey lack of consent. Validity was assessed by two criteria: (1) prevalence rates and (2) convergent and discriminant validity.3 The first research question (RQ1) addressed by this study is: Does the language used to convey lack of consent affect participants’ responses? There was no strong theoretical rationale or empirical evidence to support specific hypotheses; although we could identify reasons that each phrase might be confusing to some participants. “Without her consent” corresponds to common legal language and sounds unequivocal. Nonetheless, we had two concerns. First, as used in the 2007 version of the SES, this phrase does not provide any definition or examples of what constitutes lack of consent and participants’ personal definitions of consent may not correspond to researchers’ or legal definitions. Second, due to recent media attention on the importance of obtaining consent, simply seeing the phrase “without consent” may have deterred some participants from honest reporting (Buday & Peterson, 2015; Strang & Peterson, 2017). The “make her” language used in Abbey et al. (2005) was interpreted by most of the authors as implying that some type of strong verbal or physical force was used and that participants had to be aware of their own agency to endorse the item. However, one author interpreted “make her” as a persuasive strategy in which one person’s words and actions genuinely “made” the other person change their mind and freely agree to the sex that occurred. The “when she didn’t want to” language used in Abbey et al. (2011) was interpreted by all the of the authors as demonstrating that participants were aware that the person was not freely choosing to engage in sexual activity. Whether that person never consented or eventually capitulated due to pressure, individuals who achieve sex this way have acted without freely given consent.

We conjecture that nuances between the meaning of “without consent,” “make her,” and “when she didn’t want to” may matter more for someone answering a question about verbally coercive tactics such as telling lies to obtain sex than for someone answering a question about rape tactics such as using physical force to obtain sex. It may be relatively easy in current American culture for a man to convince himself that telling lies or using verbal pressure to “change someone’s mind” is just part of the heterosexual seduction script and that many men use these strategies to turn a “no” into a “yes” (Endendijk et al., 2020; Sakaluk et al., 2014; Vannier & O’Sullivan, 2011). It seems much harder to convince oneself that physical force or sex with someone who is incapacitated are normative elements of the heterosexual seduction script. Although the SES does not use legalistic terms such as rape or sexual assault, most adults in the US are aware of what constitutes rape; they are likely much less aware of what constitutes verbal coercion (Haugen et al., 2018; Kahn et al., 2003; Pugh & Becker, 2018). This line of reasoning leads to the first hypothesis (H1), which is that the language used to convey lack of consent is more likely to affect rates of self-reported acts of verbal coercion than of rape. Following the lead of Abbey et al. (2021), if differences were found between versions, we intended to conduct exploratory analyses to determine which specific tactics contributed to these differences.

Sexual behaviors that are culturally undesirable are frequently underreported, including sexual aggression (King, 2022; Rutherford, 2017; Strang & Peterson, 2020); thus measures that yield higher prevalence rates are often judged as providing more accurate responses. However, construct validity is also an essential criteria when selecting a measure. The second research question (RQ2) is the same as in Abbey et al. (2021): Do these three measures have similar relationships with established convergent and discriminant validity indicators of sexual aggression? Evidence for construct validity is obtained when a measure is moderately to highly correlated with established risk factors (i.e., convergent validity indicators) and uncorrelated or weakly correlated with unrelated or different constructs (i.e., discriminant validity indicators). We used the same measures as did Abbey et al. with the exception that we added childhood victimization and witnessing parental violence as convergent validity indicators (and omitted rape myth acceptance). Convergent validity indicators were selected based on Tharp et al.’s (2013) thorough review of the literature and additional recent research (Abbey et al., 2018; Davis et al., 2018; King et al., 2019; Malamuth et al., 2021). Abbey et al. (2021) found highly similar relationships across versions of the SES between the total number of sexually aggressive acts that participants endorsed and the convergent and discriminant validity indicators. We expected to replicate those findings.

Method

Participants

A Qualtrics Panel was recruited for an online study of men’s dating and sexual experiences. Participants were required to be age 18 to 35 (M = 30.27; SD = 4.01), self-identify as a man, report that they had engaged in some form of sexual activity with a woman in their lifetime (ranging from passionate kissing to sexual intercourse), and that they had resided in the United States since at least age 13 (so that they were exposed to American cultural norms about gender roles and sexual scripts during adolescence). Among the 1,291 participants included in data analyses, 86.8% self-identified as White/European-American, 4.2% as Hispanic/Latinx, 4% as Black/African-American, 3.2% as East, South, or Southeast Asian 0.9% as Multi-ethnic, 0.4% as Native American, 0.3% as Arab/Middle Eastern, and 0.2% as other. Seventy-four percent of participants had completed college. Ninety-eight percent of participants had lived in the United States their entire life. Sixty-six percent of participants were currently married, 5% were not married but were living with a partner or engaged, 8% were in an exclusive dating relationship, and the remaining 21% were single and not dating exclusively.

Procedures

Participants were recruited through Qualtrics, which creates national panels from numerous sources including opt-in marketing research panels and social media. Participants are compensated with cash, gift certificates, vouchers, or rewards based on the source from which they were recruited. Researchers pay a set amount per completed survey and compensation is handled by Qualtrics (Qualtrics 28 Questions). The ad for this study indicated that it was an academic research study that would take 25 minutes to complete. Interested individuals then answered questions that determined their eligibility for the study using the criteria described in the previous section. Qualtrics Panel samples are typically representative of the target population and provide high quality responses (Belliveau et al., 2022; Boas et al., 2020; Wadell et al., 2022). The survey was administered through the Qualtrics portal using a university account. Eligible participants were presented with the information sheet and asked to signify agreement by proceeding with the survey. Participants completed the first several sections of the survey (demographics, life satisfaction, alcohol consumption, dating and consensual sexual experiences), then they completed the version of the sexual aggression measure to which they were randomly assigned through Qualtrics programming, and then they completed the remaining measures. Piloting occurred in December 2020 and the remainder of the data were collected between January 22 and May 3, 2021.

To ensure that participants were not upset by the questions, they were asked to rate their feelings at the end of the survey using 5-point scales with response options that ranged from 1 (not at all) to 5 (very). Participants reported low levels of embarrassment (M = 1.93; SD = 0.98), anger (M = 1.67; SD = 1.02), sadness (M = 2.01; SD = 1.10), and tiredness (M = 2.28; SD = 1.24). The study’s procedures were approved by the University’s Institutional Review Board.

Measures

Sexual Aggression.

The Abbey et al. (2005) modified version of the SES has nine tactics; four verbal tactics (repeated verbal pressure or persistent arguments; showed your displeasure or anger; used flattery or made promises that were probably untrue; verbal threats for example saying you will end things or spread rumors), three incapacitation tactics (put alcohol in nonalcoholic drink without their knowledge; put drugs in drink or food without their knowledge; taken advantage when passed out or too drunk or high to stop what was happening), and two physical tactics (threatened physical harm; use of force for example holding down or physically hurting) fully crossed with five sexual activities [(1) kiss or touch or rub up against in a sexual way; (2) oral sex; (3) vaginal sex; (4) anal sex or penetration with an object; (5) attempted vaginal, oral, or anal sex but it did not happen] for a total of 45 items. Participants are asked about experiences since age 14, using 4-point scales with options ranging from 0 (never) to 3 (three or more times). The three updated language versions examined in this paper differed only in the phrasing used to convey lack of consent: “make her;” “without her consent;” or “when she didn’t want to.” For example, participants read the following question with one of the three bracketed phrases depending on the condition to which they were randomly assigned, “Have you used some type of physical force with a woman, for example by holding her down or keeping her from leaving or physically hurting her, in order to [make her have vaginal sex/have vaginal sex without her consent/have vaginal sex when she didn’t want to]?

Hostile Masculinity.

The 10-item Hostility toward Women and the 15-item Adversarial Sexual Beliefs scales (Lonsway & Fitzgerald, 1995) were used to assess hostile masculinity, a multidimensional construct that includes wanting to control and dominate women, feeling hostile and distrustful of them, and viewing them as adversaries (Malamuth et al., 1995). Both of these measures are frequently used to predict sexual aggression and have demonstrated strong convergent validity and reliability (Gerger et al., 2007; Malamuth et al., 1991; Payne et al., 1999). Responses were made on 7-point scales with options ranging from 1 (strongly disagree) to 7 (strongly agree). Cronbach’s α was .84 and .88, respectively in this study.

Childhood Experiences of Violence.

Childhood experiences with violence were assessed with two measures. An abbreviated version of Bremner et al.’s (2007) Early Trauma Inventory was used to assess Childhood Physical and Emotional Abuse from Parents. This 9-item measure demonstrated good internal consistency reliability and was significantly associated with self-reported sexual aggression in Abbey et al. (2011). Responses were made on 6-point scales with options ranging from 0 (never) to 5 (daily). Based on Bremner et al.’s recommendations, the number of events that occurred at least once were summed. Cronbach’s α was .85 in this study.

Becker et al.’s (2010) 4-item measure of Witnessing Father’s Violence toward Mother was also included. This measure demonstrated good internal consistency reliability in Becker et al. (2010) and was significantly associated with posttraumatic stress symptoms. Responses were made on 5-point scales with response options ranging from 1 (never) to 5 (more than 20 times). Cronbach’s α was .89 in this study.

Impersonal Sexual Orientation.

Three measures assessed impersonal sexual orientation, which is a “noncommittal, game-playing orientation to sexual relations” (Malamuth et al., 1995, p. 354). Participants estimated their lifetime total number of consensual sexual partners and their lifetime total number of one time only consensual sexual partners. As expected from previous literature, responses were positively skewed; thus, they were winsorized (Tabachnick & Fidell, 2007). The 10-item Permissive Attitudes subscale of the Brief Sexual Attitudes Scale was also included (Hendrick et al., 2006). Responses were made on 5-point scales with options ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach’s α was .87 in this study. These three measures are often used to predict sexual aggression and have demonstrated strong convergent and discriminant validity (Abbey et al., 2011; Hendrick et al., 2006; Malamuth et al., 1991).

Friends’ Approval and Pressure for Coerced and Forced Sex.

The 6-item Friends’ Approval and Pressure for Coerced and Forced Sex measure (Abbey et al. 2001) was included to assess implicit pressure to use coercive tactics to obtain sex. Participants answered three questions that asked the extent to which their friends would approve of lying, getting a woman drunk, and using force to get a woman to have sex with them. They were then asked how much pressure they felt from their friends to engage in each of these behaviors. This measure has been previously used to predict sexual aggression and has demonstrated good convergent validity and reliability (Abbey et al., 2006; Jacques-Tiura et al., 2015; Thompson et al., 2013). Responses were made on 5-point scales with options ranging from 1 (not at all) to 5 (very much). Cronbach’s α was .90 in this study.

Alcohol Consumption.

Three measures of alcohol consumption were included. Heavy episodic alcohol consumption was assessed by asking participants how often they consumed 5 or more drinks containing alcohol in a two-hour period in the past 12 months (National Institute on Alcohol Abuse and Alcoholism, 2003). Response options ranged from 0 (0 days in the past year) to 9 (every day). Participants were also asked how much alcohol they typically consumed in consensual sexual situations and how much their partners typically consumed in consensual sexual situations (Abbey et al., 1998; 2011). Response options ranged from 0 (none) to 7 (thirteen or more drinks). Both general heavy drinking and drinking in sexual situations are positively associated with sexual aggression (Abbey et al., 2011; Testa & Cleveland, 2017).

Social Desirability.

The 20-item Impression Management subscale from the Balanced Inventory of Desirable Responding (Paulhus, 1991) assessed social desirability. Researchers have found modest correlations between social desirability and sexual aggression, which are usually nonsignificant (Abbey et al., 2018; 2021; Tharp et al., 2013). Responses are made on 7-point scales with options ranging from 1 (not true) to 7 (very true). This measure has demonstrated good validity and reliability (Kroner & Weekes, 1996; Paulhus, 1991). Cronbach’s α was .78 in this study.

Life Satisfaction.

A 7-item abbreviated version of Andrews and Robinson’s (1991) Quality of Life measure was included as a discriminant validity indicator. Both the original and modified versions of this measure exhibit high internal consistency and moderate test–retest reliability (Andrews & Robinson, 1991). Responses were made on 7-point response scales with options ranging from 1 (terrible) to 7 (delighted). Cronbach’s α was .87 in this study.

Perceived Stress.

The 4-item short form of the Perceived Stress Scale was included (Warttig et al., 2013) as a discriminant validity indicator. This measure has previously demonstrated good reliability and validity (Warttig et al., 2013). Responses were made on a 5-point scale with options ranging from 0 (never) to 4 (very often). Cronbach’s α was .66 in this study.

Love and Commitment Motives for Sex.

The 13-item Emotional Love and Commitment subscale of the YSEX Scale (Meston & Buss, 2007) was also included, as a discriminant validity indicator. This measure has demonstrated good validity and reliability (Meston & Buss, 2007). Responses were on 5-point scales with options ranging from 1 (none of my sexual experiences) to 5 (all of my sexual experiences) occurred for this reason. Cronbach’s α was .93 in this study.

Demographics.

Participants reported their gender (man; woman; do not identify as either, please specify), age, ethnicity, length of residency in the United States, relationship status, education, occupation, and household income.

Data Analysis Plan

Analyses were conducted with SPSS. One-way ANOVAs and contingency table analyses were conducted to verify that there were no differences in background characteristics between participants randomly assigned to each of the four versions of the sexual aggression measure (Updated “make her” version, n = 322; Updated “without her consent version, n = 309; Updated “when she didn’t want to” version, n = 329; Original “make her” version that is only discussed in the Supplemental Materials, n = 331). As expected, participants randomly assigned to the different versions of the sexual aggression measure did not differ in age (df = 3, 1290, F = 0.49, p = .69), education (df = 3, 1290, F = 0.56, p = .64), income (df = 3, 1290, F = 1.27, p = .28), occupation (df = 30, χ2 = 35.94, p = .21), years living in the United States (df = 3, 1290, F = 1.17, p = .32), ethnicity (df = 27, χ2 = 17.06, p = .93), or relationship status (df = 18, χ2 = 14.68, p = .68).

Although this study is not a direct replication of Abbey et al. (2021), it has similar goals and used similar measures. Thus, where relevant, we conducted the same data analyses as they did to evaluate RQ#1 and #2. RQ#1 was assessed using contingency table analyses, followed up with Bonferroni adjusted z-scores if the omnibus test was significant (Franke et al., 2012). SPSS also provided standardized residuals, which divide the residual for a cell (expected - observed) by the expected sample size for that cell. Large residuals are an indication that cell contributed to the overall significant difference.

There are multiple ways to score the SES and researchers have formed different variables based on their research aims (Davis et al., 2014). As in Abbey et al. (2021), two perpetration variables were used in the primary analyses. First participants were coded as either perpetrators (reported committing one or more acts of any type of sexual aggression since age 14) or nonperpetrators (no reported acts of sexual aggression since age 14). Then participants were coded into one of five mutually exclusive categories based on the most severe type of sexual aggression they reported committing since age 14: nonperpetrator, nonpenetrative sexual contact, completed verbally coercive penetrative sex, attempted rape, and completed rape. This variable has been commonly used in past research and operationalizes severity in terms of the legal system’s assessment of the seriousness of the crime and by prioritizing penetrative acts (Koss et al., 1987; 2007).

If differences were found, then several exploratory analyses were planned based on those conducted by Abbey et al. (2021). First, we created three new variables: any use of verbally coercive tactics; any use of alcohol, drug, or incapacitation tactics; or any use of physical force or threats of physical force tactics.4 Unlike the most severe type of sexual aggression variable described above, these variables are not mutually exclusive. Each participant received a score of 1 if they reported using that type of tactic for any type of sexual activity at least once since age 14 or 0 if they reported no instances. Contingency table analyses would then be used to determine if the percentage of participants who reported using each of these three types of tactics differed based on the consent language. Second, if significant differences were found for one or more types of tactics, then we planned on conducting contingency table analyses at the item level to determine for which items these differences occurred.

RQ#2 was assessed with bivariate correlations between each of the sexual aggression measures and the convergent and discriminant validity indicators. The total number of sexually aggressive acts perpetrated was used as the primary outcome measure (Abbey et al., 2011; Malamuth et al., 1991). This was calculated by summing participants’ responses to each SES item. Cronbach’s alpha was .98 for each of the three versions. The SES does not include questions that allows researchers to determine the number of separate incidents that occurred because multiple tactics and sex acts could have been used in a single incident. We used an online calculator (Lenhard & Lenhard, 2014) to determine if the magnitude of the correlations between a validity indicator and sexual aggression significantly differed based on the language used to convey lack of consent.

Sample Size Determination.

G* Power (Faul et al., 2009) was used to determine that a sample size of 1288 was required to compare 4 versions of the sexual aggression measure with the assumptions of alpha = .05, power = .80, and a small (.10) effect size.

Data Quality Checks.

Based on commonly used guidelines for data cleaning (Curran, 2016; DeSimone, Harms, & DeSimone, 2015; Huang, Liu, & Bowling, 2015), we established data quality checks for Qualtrics to use. These included providing an incorrect response to more than one of the three attention checks, taking less than half the median time required by the first 300 participants to complete the survey (i.e., less than 9.95 minutes), not completing the survey, skipping 20% or more of the questions, providing answers to multiple open-ended questions which were not responsive, and providing long strings of the same response to close-ended questions across different measures. Based on past experience with survey participants who appeared to try to guess the inclusion criteria, the inclusion criteria were repeated within the body of the survey and individuals were eliminated who did not meet them.5 Among the 1,605 individuals who were not initially screened out of the survey, 201 quit before completing the survey, 45 provided a response to the eligibility criteria within the body of the survey that made them ineligible, 29 did not provide a correct response to at least two of the three attention checks, 24 provided answers to multiple open-ended questions that demonstrated they were not reading the questions, 12 provided long strings of the same response to close-ended questions, and 1 had 20% or more missing data (none who met all the other criteria was eliminated for completing the survey too quickly). Two surveys were removed because inspection of their open-ended responses identified them as the same individual. The resulting sample size was 1,291. Less than 1% of the data were missing; thus, mean substitution was used (Tabachnick & Fidell, 2007).

Results

Research Question #1: Does the Language Used to Convey Lack of Consent Affect Self-Reported Rates of Perpetration?

The 2 (any sexual aggression: no or yes) X 3 (lack of consent phrasing) contingency table analysis was significant, χ2(2, N = 960) = 28.62, p < .001. Although all three versions identified alarmingly high self-reported rates of perpetration since age 14, participants randomly assigned to the “make her” language reported significantly higher rates of perpetration (73.3% adjusted standardized residual = 2.4) than did participants randomly assigned to the “without her consent” language (52.8%; adjusted standardized residual = −2.2, p < .001) or the “when she didn’t want to” language (62.0%; adjusted standardized residual = −0.2, p = .006), and those two versions did not significantly differ from each other.

The 5 (most severe type of sexual aggression: none, nonpenetrative sexual contact, verbally coerced penetrative sex, attempted rape, or completed rape) X 3 (lack of consent phrasing) contingency table analysis was also significant, χ2(8, N = 960) = 58.33, p < .001. As can be seen in Table 1, a significantly higher percentage of participants randomly assigned to the “make her” language reported verbally coerced penetrative sex as their most severe act of sexual aggression as compared to those randomly assigned to the “without her consent” and “when she didn’t want to” language. No significant differences were found in the percentage of participants who reported nonpenetrative sexual contact, attempted rape, or completed rape as their most severe act of sexual aggression (see Table 1). Thus, as hypothesized, the language used to convey lack of consent had significant effects on self-reports of verbally coerced penetrative sex; however, there were no significant effects on self-reports of completed rape (or nonpenetrative sexual contact or attempted rape).

Table 1.

Comparison of Men’s Self-Reported Most Severe Type of Sexual Aggression Since Age 14 Using Three Different Phrases to Convey Lack of Consent (N = 960)

“Make Her” (n = 322) “Without Her Consent” (n = 309) “When She Didn’t Want To” (n = 329)
Perpetration Group1 % n Standardized Residual % n Standardized Residual % n Standardized Residual
Nonperpetrators 26.7 a 86 −3.1 47.2 b 146 2.9 38.0 b 125 0.2
Sexual Contact 5.6 a 18 −1.9 10.4 a 32 1.0 10.0 a 33 0.9
Verbal Coercion 32.0 a 103 4.4 13.6 b 42 −2.8 16.7 b 55 −1.6
Attempted Rape 0.3 a 1 −1.0 1.3 a 4 0.9 0.9 a 3 0.2
Completed Rape 35.4 a 114 .9 27.5 a 85 −1.5 34.3 a 113 0.6
Open in a new tab
1

Groups are mutually exclusive so each column totals 100%.

Note. Values with different subscripts in the same row are significantly different from each other at p < .05 using a Bonferroni adjustment.

Follow-up exploratory analyses.

We followed up on these findings in several ways. The percentage of participants who endorsed any verbally coercive tactic significantly differed across the three versions; with the highest rates reported for the “make her” language (71.4%), the second highest rate for the “when she didn’t want to” language (58.4%), and the lowest rate for the “without her consent” language (47.9%), with all three versions significantly different from each other, χ2(2, N = 960) = 36.42, p <.001.6 There were no significant differences between versions in endorsement of any alcohol, drug, or unable to consent due to incapacitation tactics; or any physical force or threats of force tactics, χ2(2, N = 960) = 1.86, p = .40; 1.29, p = .52, respectively.

Because differences were only found for verbal coercion tactics, we examined participants’ endorsement of each of the 4 verbal coercion items to determine if the pattern of differences associated with the language used to convey lack of consent was similar for all of them. As can be seen in Table 2, regardless of the language used to convey lack of consent, pressure and flattery were most commonly endorsed and verbal threats were least commonly endorsed. Participants randomly assigned to the “without her consent” language had lower rates of endorsement of all four verbal coercion tactics as compared to participants randomly assigned to the “make her” or the “when she didn’t want to” language. For verbal pressure and flattery, “when she didn’t want to” produced significantly lower rates of endorsement than did “make her”; whereas, for showed displeasure and verbal threats, rates of endorsement did not significantly differ for “when she didn’t want to” and “make her.”

Table 2.

Correlations between Convergent and Discriminant Validity Indicators and Sexual Aggression Using Three Different Phrases to Convey Lack of Consent (N = 960)

“Make her”

(n =322 )		 “Without her consent”
(n = 309)		 “When she didn’t want to”
(n = 329)
Convergent Validity Indicators:
 Hostility toward Women .47** .42** .46**
 Adversarial Sexual Beliefs .49** .42** .46**
 Childhood Victimization1 .24**a .29** .41**b
 Witness Father Violence2 .20**a .41**b .37**b
 Total No. Sex Partners .06 −.05 .07
 Total No. One Time Only Sex Partners .06 .07 .17**
 Positive Attitudes about Casual Sex .30** .31** .28**
 Friends’ Approval Coerced Sex .45**a .48**a .63**b
 Heavy Episodic Alcohol Consumption .41** .33** .39**
 Own Alcohol Sexual Situations3 .28** .35** .32**
 Partner Alcohol Sexual Situations4 .25** .29** .36**
Discriminant Validity Indicators:
 Social Desirability .10a .09a .25**b
 Life Satisfaction .20**a .03b .11
 Perceived Stress .04 .15** .16**
 Love and Commitment Motives for Sex −.02 −.04 .09
Open in a new tab

Note. Correlations in the same row with different subscripts are significantly different from each other, p < .05.

1

Childhood Emotional and Physical Abuse by Parents or Caregivers

2

Witnessing Father’s Violence Toward Mother

3

Typical Quantity of Alcohol Consumed in Consensual Sexual Situations with Women

4

Typical Quantity of Alcohol Consumed by Woman Partner in Consensual Sexual Situations

**

p < .01.

Research Question #2: Does the Language used to Convey Lack of Consent Affect Convergent and Discriminant Validity?

Table 2 presents correlations that assess the second research question. The pattern of results for the 11 convergent validity measures was highly comparable for all three phrases used to convey lack of consent. The greater the number of sexually aggressive acts that participants reported perpetrating, the greater their hostility toward women, endorsement of adversarial sexual beliefs, experience of childhood emotional and physical abuse by a parent or caregiver, witnessing their father being violent toward their mother, positive attitudes toward casual sexual experiences, their friend’s approval of using coercive strategies to obtain sex, their heavy episodic alcohol consumption, their usual alcohol consumption in sexual situations, and their partners’ usual alcohol consumption in sexual situations. As can be seen in Table 2, the magnitude of some of the correlations significantly differed; however, the patterns were similar. The only common risk factors that were not significantly correlated with perpetration (with one exception) were the total number of (lifetime) sexual partners and the total number of one time only sex partners. The exception was a modest significant correlation between lifetime number of partners and perpetration with the “when she didn’t want to” language.

Similarly, there were comparable small correlations between the number of sexually aggressive acts that participants reported perpetrating and social desirability, life satisfaction, perceived stress, and love and commitment motives for having sex (see Table 2). The correlation between social desirability and number of sexually aggressive acts was not significantly different from zero for “make her” and “without her consent”; although it was modestly positively correlated with “when she didn’t want to” (and significantly different from the other two).

Discussion

The major goal of this study was to compare the effects of three different phrases used to convey lack of consent on men’s self-reports of sexual aggression toward women. Random assignment to measures that were identical except for the language used to convey lack of consent ensures that any differences found are due to those differences in phrasing. The answer to the first research question is that the language used to convey lack of consent significantly affected reports of verbal coercion only. Two different analyses support this conclusion. First, when mutually exclusive groups were formed, significantly more participants reported that verbally coerced penetrative sex was their most severe type of perpetration when lack of consent was operationalized as “make her” as compared to when it was operationalized as “without her consent” or “when she didn’t want to.” In contrast, there were no significant differences in participants’ reports of completed rape, attempted rape, or nonpenetrative sexual contact as their most severe type of perpetration. Second, the percentage of participants who reported that they had used a verbally coercive tactic at least once (regardless of the most severe type of sexual aggression they committed) was significantly highest for the “make her” language and lowest for the “without her consent language” with “when she didn’t want to” falling in between and significantly different from both. In contrast, there were no significant differences in rates of endorsement for either of the types of tactics that are associated with rape: incapacitation and physical force.

These findings support Hypothesis 1 and provide a conceptual replication of Abbey et al.’s (2021) experiment, which also found significant differences in rates of verbal coercion when comparing three different sexual aggression measures. These findings suggest that the language used to convey lack of consent may not have much impact on participants’ interpretation of questions about the use of physical force, threats of physical force, or sex when someone is too incapacitated to stop what is happening due to alcohol or drugs or another reason. The word “rape” does not have to be included in the question for participants to recognize that these acts are not socially acceptable. In contrast, there appears to be more ambiguity regarding the acceptability of verbally coercive tactics, presumably because traditional sexual scripts normalize the use of lies, false promises, flattery, nagging, arguments, and verbal threats to “convince” someone to have sex. Thus with these tactics, some participants may be less sure as to whether their past behavior fits the question and parse the language to determine how they should respond. Someone might think to themselves “I would never do that without consent; I know that’s wrong. But it’s okay to make her want to have sex with me by promising to spend more time with her, even if I’m not sure that I will.” This example is purely hypothetical and qualitative research is needed that carefully examines how participants interpret common terms used to convey that sex occurred without consent.

Despite these differences in verbal coercion prevalence, the answer to the second research question is that all three measures demonstrated good convergent validity. For each measure, the total number of reported sexually aggressive acts was moderately highly correlated with well-known risk factors for sexual aggression (Tharp et al., 2013). Additionally, each of the three measures demonstrated good discriminant validity by having much lower correlations with measures not typically associated with sexual aggression. Given that all these phrases are used to convey lack of consent in existing measures, these findings are reassuring.

Despite the comments above regarding how some language may deter reporting, rates of self-reported sexual aggression were extremely high in this study. Depending on the language used to convey lack of consent, 52.8% to 73.3% of this community-based sample reported at least one act of sexual aggression since age 14. Rates of completed rape ranged from 27.5% to 35.4%. Many past studies have relied on college samples, comprised primarily of single men between the ages of 18 and 22. Participants in this study were 30 years of age on average and two-thirds were married. Many of these participants could have been describing marital sexual aggression, which is not typically assessed with the SES. These findings need to be replicated; however, they highlight the importance of addressing sexual aggression in the general community, not just on college campuses or only with young, single men.

Limitations

Participants were mostly white, college educated, and married. The limited ethnic diversity was unexpected because Qualtrics Panels are typically representative of the sampled segment of the US population (Belliveau et al., 2022; Boas et al., 2020). Given this study’s focus on explaining men’s sexual aggression toward women, it was limited to individuals who identified as a man and had engaged in sexual activities with women. Although we included a number of attention and quality checks, participants who complete online surveys may be not as attentive as participants who complete in person surveys. Participants who complete surveys online for compensation typically do not provide detailed responses to open-ended questions; thus, this methodology was not optimal for evaluating participants’ perceptions of the meaning of each of the three phrases used to convey lack of consent.

Future Research Directions

The findings from this study provide a “yes” answer to the question, “Does the language used to convey lack of consent affect participants’ endorsement of sexual aggression items?” A logical follow-up question is “Which language provides the most accurate responses?’ Answering this question requires asking participants to provide more detailed information about the incident that goes beyond standard measures designed to assess prevalence. Instruments like the SES need to balance the competing goals of being detailed enough to elicit the full range of behaviors being assessed yet brief enough that they will be adopted by other researchers and administered widely. Determining if sexual activity occurred without freely given consent typically requires detailed information about the incident (Littleton et al., 2019; Testa et al., 2004). Recent studies have found that some participants (particularly women) misunderstand questions asked in sexual aggression surveys and report on situations in which they were the victim, not the perpetrator (Buday & Peterson, 2013; Jeffrey & Senn, 2024; Strang & Peterson, 2017). These same studies have also found misunderstandings in the opposite direction, with some participants labeling some situations as consensual although consent was not freely obtained. It is unlikely that everyone interprets the same phrasing in precisely the same way, particularly if studies are conducted with individuals who are diverse in ethnicity, culture, religion, gender, and sexual orientation. Definitions of key terms can help mitigate these types of errors, although research is needed to determine if most participants actually read the instructions and recall them as they answer the questions.

Prevention and Policy Implications

Given that more than half of the participants in this community sample acknowledged that they committed at least one sexually aggressive act since age 14, preventionists and policy makers need to focus on initiatives at the societal level. A sizable proportion of these acts involved the use of physical force and having sex with someone too incapacitated to consent. And even higher proportion involved the use of verbally coercive tactics. Because these tactics have been normalized, particularly in traditional heterosexual relationships, perpetrators and outsiders can convince themselves that victims changed their minds, rather than recognizing that their persistence made resistance seem futile. Beginning in early childhood, everyone should learn to respect other people’s bodily integrity and that everyone has the right to make their own sexual decisions without coercion. Given the high rates of sexual assault in adolescent and adult relationships (Leemis et al., 2022), these programs need to continue into adulthood and focus on the importance of establishing freely given consent for sexual activity at that time with steady partners, as well as new partners. Social media and advocacy initiatives are needed to encourage movies (including erotica), television shows, books, and other media to accurately represent sexual consent and the negative outcomes of coerced sex. It is not surprising that youth develop sexual scripts that include coercion if society consistently provides exemplars that are rewarded for engaging in this behavior.

Supplementary Material

supplement
2

Acknowledgments

This research was supported by funding from the Wayne State University Psychology Department to Antonia Abbey. We have no known conflict of interest to disclose.

Footnotes

1

A new version of the SES victimization version is in press (Koss et al., 2024). The perpetration version has not been revised as of yet and this study was conducted before the new victimization version was under development.

2

Terms used in Abbey et al., (2005) to describe some tactics and sex acts were updated by the research team before conducting this study. To ensure that the updated language did not substantially affect participants’ responses, the original Abbey et al. (2005) measure was included as a fourth condition to which participants were randomly assigned. This allowed the two versions that used the “make her” language to be compared. The findings regarding this secondary research question are described in Section 1 of the Supplementary Materials.

3

Participants were also asked an open-ended question to provide preliminary information about how they interpreted each of the phrases used to convey lack of consent. The authors coded the substantive responses for exploratory purposes. A summary is included in Section 2 of the Supplementary Materials.

4

Abbey et al. (2021) only report on verbal coercion because that is where differences were found. Based on a reviewer’s suggestion, we examined all three types of tactics.

5

Qualtrics made an error in their initial programming that eliminated 35 year old men. This error was corrected (so some 35 year olds were included in the dataset) but then reappeared in later releases of the survey.

6

These rates are higher than those for the most severe verbally coercive penetrative sex group reported in Table 1 because these rates include everyone who reported using a verbally coercive tactic, regardless of whether it was used for penetrative sex or nonpenetrative sex and regardless of whether the participant also reported an act of attempted or completed rape.

References

  1. Abbey A, Helmers BR, Jilani Z, McDaniel MC, & Benbouriche M (2021). Assessment of men’s sexual aggression against women: An experimental comparison of three versions of the sexual experiences survey. Psychology of Violence, 11, 253–263. [Google Scholar]
  2. Abbey A, Jacques-Tiura AJ, & LeBreton J (2011). Risk factors for sexual aggression in young men: An expansion of the confluence model. Aggressive Behavior, 37, 450–464. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Abbey A, McAuslan P, & Ross LT (1998). Sexual assault perpetration by college men: The role of alcohol, misperception of sexual intent, and sexual beliefs and experiences. Journal of Social and Clinical Psychology, 17, 167–195. [Google Scholar]
  4. Abbey A, Parkhill MR, BeShears R, Clinton-Sherrod AM, & Zawacki T (2006). Cross-sectional predictors of sexual assault perpetration in a community sample of single African American and Caucasian men. Aggressive Behavior, 32, 54–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Abbey A, Parkhill MR, & Koss MP (2005). The effects of frame of reference on responses to questions about sexual assault victimization and perpetration. Psychology of Women Quarterly, 29, 364–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Abbey A, Pegram SE, Woerner J, & Wegner R (2018). Men’s responses to women’s sexual refusals: Development and construct validity of a virtual dating simulation of sexual aggression. Psychology of Violence, 8, 87–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Adams-Curtis LE, & Forbes GB (2004). College women’s experiences of sexual coercion: a review of cultural, perpetrator, victim, and situational variables. Trauma, violence & abuse, 5(2), 91–122. [Google Scholar]
  8. Anderson RE, Cahill SP, & Delahanty DL (2017). Initial evidence for the reliability and validity of the Sexual Experience Survey-Short Form Perpetration (SES-SFP) in college men. Journal of Aggression, Maltreatment and Trauma, 26, 626–643. [Google Scholar]
  9. Andrews FM, & Robinson JP (1991). Measures of subjective well-being. In Robinson JP, Shaver PR, & Wrightsman LS (Eds.). Measures of personality and social psychological attitudes (pp. 61–114). San Diego: Academic Press. [Google Scholar]
  10. Becker KD, Stuewig J, & McCloskey LA (2010). Traumatic stress symptoms of women exposed to different forms of childhood victimization and intimate partner violence, Journal of Interpersonal Violence, 25, 1699–1715. [DOI] [PubMed] [Google Scholar]
  11. Belliveau J, Soucy KI, & Yakovenko I (2022). The validity of Qualtrics panel data for research on video gaming and gaming disorder. Experimental and Clinical Psychopharmacology, 30(4), 424–431. [DOI] [PubMed] [Google Scholar]
  12. Boas T, Christenson D, & Glick D (2020). Recruiting large online samples in the United States and India: Facebook, Mechanical Turk, and Qualtrics. Political Science Research and Methods, 8(2), 232–250. [Google Scholar]
  13. Bremner JD, Bolus R, & Mayer EA (2007). Psychometric properties of the Early Trauma Inventory-Self-Report. Journal of Nervous and Mental Disease, 195, 211–218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Buday SK, & Peterson ZD (2015). Men’s and women’s interpretation and endorsement of items measuring self-reported heterosexual aggression. Journal of Sex Research, 52, 1042–1053. [DOI] [PubMed] [Google Scholar]
  15. Cook SL (2002). Self-reports of sexual, physical, and nonphysical abuse perpetration. Violence Against Women, 8, 541–565. [Google Scholar]
  16. Cook SL, Gidycz CA, Koss MP, & Murphy M (2011). Emerging issues in the measurement of rape victimization. Violence Against Women, 17(2), 201–218. [DOI] [PubMed] [Google Scholar]
  17. Curran PG (2016). Methods for the detection of carelessly invalid response in survey data. Journal of Experimental Social Psychology, 66, 4–19. [Google Scholar]
  18. Davis KC, Gilmore AK, Stappenbeck CA, Balsan MJ, George WH, & Norris J (2014). How to score the Sexual Experiences Survey? A comparison of nine methods. Psychology of Violence, 4(4), 445–461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Davis KC, Masters NT, Casey E, Kajumulo KF, Norris J, & George WH (2018). How childhood maltreatment profiles of male victims predict adult perpetration and psychosocial functioning. Journal of Interpersonal Violence, 33(6), 915–937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. De Coster S, & Heimer K (2021). Unifying theory and research on intimate partner violence: A feminist perspective. Feminist Criminology, 16(3), 286–303. [Google Scholar]
  21. DeSimone JA, Harms PD, & DeSimone AJ (2015). Best practice recommendations for data screening. Journal of Organizational Behavior, 36(2), 171–181. [Google Scholar]
  22. Endendijk JJ, van Baar AL, & Deković M (2020). He is a stud, She is a slut! A meta-analysis on the continued existence of sexual double standards. Personality and Social Psychology Review, 24(2), 163–190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Faul F, Erdfelder E, Buchner A, & Lang AG (2009). Statistical power analyses using G* Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149–1160. [DOI] [PubMed] [Google Scholar]
  24. Franke T , Ho T, & Christie CA (2012). The chi-square test: Often used and more often misinterpreted. American Journal of Evaluation, 33, 448–458. [Google Scholar]
  25. Gerger H, Kley H, Bohner G, & Siebler F (2007). The acceptance of modern myths about sexual aggression scale: Development and validation in German and English. Aggressive Behavior, 33, 422–440. [DOI] [PubMed] [Google Scholar]
  26. Haugen AD, Rieck SM, Salter PS, & Phillips NL (2018). What Makes It Rape? A Lay Theories Approach to Defining Rape Among College Students. Basic and Applied Social Psychology, 40(1), 18–35. [Google Scholar]
  27. Hendrick C, Hendrick SS, & Reich DA (2006). The brief sexual attitudes scale. The Journal of Sex Research, 43, 76–86. [DOI] [PubMed] [Google Scholar]
  28. Huang JL, Liu M, & Bowling NA (2015). Insufficient effort responding: Examining an insidious confound in survey data. Journal of Applied Psychology, 100, 828–845. [DOI] [PubMed] [Google Scholar]
  29. Hunnicutt G (2009). Varieties of patriarchy and violence against women: Resurrecting “patriarchy” as a theoretical tool. Violence Against Women, 15(5), 553–573. [DOI] [PubMed] [Google Scholar]
  30. Jacques-Tiura AJ, Abbey A, Wegner R, Pierce J, Pegram SE, & Woerner J (2015). Friends matter: Protective and harmful aspects of male friendships associated with past year sexual aggression in a community sample of young men. American Journal of Public Health, 105, 1001–1007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jeffrey NK, & Senn CY (online first 2024). Gender differences in sexual violence perpetration behaviors and validity of perpetration reports: A mixed-method study, The Journal of Sex Research. [Google Scholar]
  32. Johnson SM, Murphy MJ, & Gidycz CA (2017). Reliability and validity of the Sexual Experiences Survey-Short Forms. Violence and Victims, 32, 78–92. [DOI] [PubMed] [Google Scholar]
  33. Jozkowski KN, Peterson ZD, Sanders SA, Dennis B, & Reece M (2014). Gender differences in heterosexual college students’ conceptualizations and indicators of sexual consent: Implications for contemporary sexual assault prevention education. Journal of Sex Research, 51(8), 904–916. [DOI] [PubMed] [Google Scholar]
  34. Kahn AS, Jackson J, Kully C, Badger K, & Halvorsen J (2003). Calling it Rape: Differences in Experiences of Women Who do or do not Label their Sexual Assault as Rape. Psychology of Women Quarterly, 27(3), 233–242. [Google Scholar]
  35. Kanin EJ, (1967). An examination of sexual aggression as a response to sexual frustration. Journal of Marriage and Family, 29,(3), 428–433. [Google Scholar]
  36. King BM (2022). The influence of social desirability on sexual behavior surveys: A review. jozArchives of Sexual Behavior, 51, 1495–1501. [Google Scholar]
  37. King AR, Kuhn SK, Strege C, Russell TD, & Kolander T (2019). Revisiting the link between childhood sexual abuse and adult sexual aggression. Child Abuse & Neglect, 94, 104022. [DOI] [PubMed] [Google Scholar]
  38. Koss MP, Abbey A, Campbell R, Cook S, Norris J, Testa M, Ullman S, West C, & White J (2007). Revising the SES: A collaborative process to improve assessment of sexual aggression and victimization. Psychology of Women Quarterly, 31, 357–370. [Google Scholar]
  39. Koss MP, Gidycz CA, & Wisniewski N (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55, 162–170. [DOI] [PubMed] [Google Scholar]
  40. Koss MP, & Oros CJ (1982). Sexual Experiences Survey: A research instrument investigating sexual aggression and victimization. Journal of Consulting and Clinical Psychology, 50, 455–457. [DOI] [PubMed] [Google Scholar]
  41. Koss MP, Anderson RA, Peterson Z, Littleton H, Abbey A, Kowalski R, Thompson M, Canan S. White, McCauley JW, H., Orchowski L, Fedina L, Lopez E, Allen C, & Swartout K (in press). The revised victimization version of the Sexual Experiences Survey (SES-V): Process, foundations, evidence, and items. Journal of Sex Research. [Google Scholar]
  42. Krahe B, and Berger A (2013). 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. Aggressive Behavior, 39(5), 391–404. [DOI] [PubMed] [Google Scholar]
  43. Kroner DG, & Weekes JR (996). Balanced Inventory of Desirable Responding: Factor structure, reliability, and validity with an offender sample. Personality and Individual Differences, 21, 323–333. [Google Scholar]
  44. Leemis RW, Friar N, Khatiwada S, Chen MS, Kresnow M, Smith SG, Caslin S, & Basile KC (2022). The National Intimate Partner and Sexual Violence Survey: 2016/2017 Report on Intimate Partner Violence. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. [Google Scholar]
  45. Lenhard W & Lenhard A (2014). Hypothesis tests for comparing correlations. available: https://www.psychometrica.de/correlation.html.Psychometrica.
  46. Littleton H, Layh M, Rudolph K, & Haney L (2019). Evaluation of the Sexual Experiences Survey—Revised as a screening measure for sexual assault victimization among college students. Psychology of Violence, 9(5), 555–563. [Google Scholar]
  47. Lonsway KA, & Fitzgerald L F. (1995). Attitudinal antecedents of rape myth acceptance: A theoretical and empirical reexamination. Journal of Personality and Social Psychology, 68, 704–711. [Google Scholar]
  48. Malamuth NM, Lamade RV, Koss MP, Lopez E, Seaman C, & Prentky R (2021). Factors predictive of sexual violence: Testing the four pillars of the Confluence Model in a large diverse sample of college men. Aggressive Behavior, 47(4), 405–420. [DOI] [PubMed] [Google Scholar]
  49. Malamuth NM, Linz D, Heavey CL, Barnes G, & Acker M (1995). Using the confluence model of sexual aggression to predict men’s conflict with women: a 10-year follow-up study. Journal of Personality and Social Psychology, 69(2), 353–369. [DOI] [PubMed] [Google Scholar]
  50. Malamuth NM, Sockloskie RJ, Koss MP, & Tanaka JS (1991). Characteristics of aggressors against women: Testing a model using a national sample of college students. Journal of Consulting and Clinical Psychology, 59, 670–681. [DOI] [PubMed] [Google Scholar]
  51. Merriam-Webster (n.d.) Merriam-Webster.com dictionary. Retrieved June 25, 2022 from https://www.merriam-webster.com/dictionary/
  52. Meston CM, & Buss DM (2007). Why humans have sex. Archives of Sexual Behavior, 36, 477–507. [DOI] [PubMed] [Google Scholar]
  53. Muehlenhard CL, Humphreys TP, Jozkowski KN, & Peterson ZD (2016) The complexities of sexual consent Among college students: A conceptual and empirical review, The Journal of Sex Research, 53(4), 457–487. [DOI] [PubMed] [Google Scholar]
  54. National Institute on Alcohol Abuse and Alcoholism (2003). Task Force on Recommended Alcohol Questions. Retrieved January 15, 2018 from https://www.niaaa.nih.gov/research/guidelines-and-resources/recommended-alcohol-questions.
  55. Ouimette PC, Shaw J, Drozd JF, & Leader J (2000). Consistency of reports of rape behaviors among nonincarcerated men. Psychology of Men & Masculinity, 1, 133–139. [Google Scholar]
  56. Paulhus DL (1991). Balanced Inventory of Desirable Responding (BIDR). In Robinson JP, Shaver PR, & Wrightsman LS (Eds.), Measures of personality and social psychological attitudes (pp. 37–41). San Diego, CA: Academic Press. [Google Scholar]
  57. Payne DL, Lonsway KA, & Fitzgerald LF (1999). Rape myth acceptance: Exploration of its structure and its measurement using the Illinois Rape Myth Acceptance Scale. Journal of Research in Personality, 33, 27–68. [Google Scholar]
  58. Pugh B, & Becker P (2018). Exploring definitions and prevalence of verbal sexual coercion and its relationship to consent to unwanted sex: Implications for affirmative consent standards on college campuses. Behavioral Sciences, 8(8), 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Qualtrics, 28 questions to help buyers of online samples. https://www.iup.edu › files › qualtrics › esomar Downloaded November 12, 2022.
  60. Reich CM, Anderson GD, & Maclin R (2022). Why I didn’t report: Reasons for not reporting sexual violence as stated on Twitter. Journal of Aggression, Maltreatment & Trauma, 31, (4), 478–496. [Google Scholar]
  61. Rutherford A (2017). Surveying rape: Feminist social science and the ontological politics of sexual assault. History of the Human Sciences, 30(4), 100–123. [Google Scholar]
  62. Sakaluk JK, Todd LM, Milhausen R, Lachowsky NJ, & Undergraduate Research Group in Sexuality (2014). Dominant heterosexual sexual scripts in emerging adulthood. Journal of Sex Research, 51, 516–531. [DOI] [PubMed] [Google Scholar]
  63. Salazar LF, Vivolo-Kantor A, & McGroarty-Koon K (2017). Formative research With college men to inform content and messages for a web-based sexual violence prevention program. Health Communication, 32(9), 1133–1141. [DOI] [PubMed] [Google Scholar]
  64. Schobert K, Cooper S, Fries N, & Sonia Chervillil M (2021). I thought we were vibin’: A qualitative exploration of sexual agency and consent in young people. Sexualities, 24(7), 906–921. [Google Scholar]
  65. Shumlich EJ, & Fisher WA (2020). An Exploration of Factors That Influence Enactment of Affirmative Consent Behaviors. The Journal of Sex Research, 57(9), 1108–1121. [DOI] [PubMed] [Google Scholar]
  66. Spohn C (2020). Sexual assault case processing: The more things change, the more they stay the same. The International Journal for Crime, Justice and Social Democracy, 9(1), 86–94. [Google Scholar]
  67. Strang E, & Peterson ZD (2020). Use of a bogus pipeline to detect men’s underreporting of sexually aggressive behavior. Journal of Interpersonal Violence, 35(1–2), 208–232. [DOI] [PubMed] [Google Scholar]
  68. Strang E, & Peterson ZD (2017). Unintentional misreporting on self-report measures of sexually aggressive behavior: An interview study. Journal of Sex Research, 54, 971–983. [DOI] [PubMed] [Google Scholar]
  69. Strang ET, Peterson ZD, Hill YN, & Heiman JR (2013). Discrepant responding across self-report measures of men’s coercive and aggressive sexual strategies. Journal of Sex Research, 50, 458–469. [DOI] [PubMed] [Google Scholar]
  70. Tabachnick BG, & Fidell LS (2007). Using multivariate statistics (5th ed.). Allyn & Bacon. [Google Scholar]
  71. Testa M, & Cleveland MJ (2017). Does alcohol contribute to college men’s sexual assault perpetration? Between- and within- person effects over five semesters. Journal of Studies on Alcohol and Drugs, 78, 5–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Testa M, VanZile-Tamsen C, Livingston JA, & Koss MP (2004). Assessing women’s experiences of sexual aggression using the Sexual Experiences Survey: Evidence for validity and implications for research. Psychology of Women Quarterly, 28(3), 256–265. [Google Scholar]
  73. Tharp AT, DeGue S, Valle LA, Brookmeyer KA, Massertti GM, & Matjasko JL (2013). A systematic qualitative review of risk and protective factors for sexual violence perpetration. Trauma, Violence, & Abuse, 14, 133–167. [Google Scholar]
  74. Thompson MP, Swartout KM, & Koss MP (2013). Trajectories and predictors of sexually aggressive behaviors during emerging adulthood. Psychology of Violence, 3, 247–259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. United Nations Human Rights Council. (2021). A framework for legislation on rape. https://digitallibrary.un.org/record/3929055
  76. Vannier SA, & O’Sullivan LF (2011.) Communicating interest in sex: Verbal and nonverbal initiation of sexual activity in young adults’ romantic dating relationships. Archives of Sexual Behavior, 40, 961–969. [DOI] [PubMed] [Google Scholar]
  77. Waddell TF, Overton H, & McKeever R (2022). Does sample source matter for theory? Testing model invariance with the influence of presumed influence model across Amazon Mechanical Turk and Qualtrics Panels. Computers in Human Behavior, 137. [Google Scholar]
  78. Warttig LS, Forshaw MJ, South J, & White AK (2013). New, normative, English-sample data for the Short Form Perceived Stress Scale. Journal of Health Psychology, 18, 1617–1628. [DOI] [PubMed] [Google Scholar]
  79. Wilson LC, & Miller KE (2016). Meta-analysis of the prevalence of unacknowledged rape. Trauma, Violence, & Abuse, 17(2), 149–159. [Google Scholar]
  80. Wolitzky-Taylor KB, Resnick HS, Amstadter AB, McCauley JL, Ruggiero JL, Kilpatrick DG (2011). Reporting rape in a national sample of college women. Journal of American College Health, 59 (7), 582–587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. World Health Organization (2013). Global and regional estimates of violence against women. Geneva, Switzerland: World Health Organization. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplement
NIHMS2056305-supplement-supplement.pdf (165.6KB, pdf)
2
NIHMS2056305-supplement-2.pdf (124KB, pdf)

RESOURCES