Abstract
Sexual assault occurs at high rates on college campuses today. This problem persists despite the many educational programs that seek to educate college students about sexual assault and consent. The current study used latent profile analysis on a survey of 717 college students at a Midwestern university to determine if different people have varying behavioral intentions about consent during four different sexual situations. Results show that there are four distinct profiles of intentions related to consent: the respecting explicit, affirmative, and ongoing consent profile (76.55% of the sample); only respecting explicit consent profile (9.89%); unsure about nonexplicit consent profile (9.04%); and the does-not-ask for verbal consent in relationships profile (4.55%). Regression analysis also revealed that males were more likely to be classified in the only respecting explicit consent profile compared with the three other profiles, but there were no significant differences among the other profiles.
Approximately 23% of female and 5% of male undergraduates reported being sexually assaulted while attending college (Cantor et al., 2015). The majority of the perpetrators were someone the student knew (e.g., a romantic partner, friend, or acquaintance; Cantor et al., 2015). Although obtaining consent seems to be straightforward, research has shown that explicit, affirmative, and mutual indications of consent are often missing in sexual encounters (Beres, 2007). Sexual relations often occur with only nonverbal consent, and individuals often prefer to assume consent rather than obtain verbal consent (Hall, 1998; Humphreys, 2007). Accordingly, it is imperative to explore how and when college students obtain and respect consent during sexual encounters.
Sexual Script Theory
Sexual script theory states that sexuality is socially constructed, meaning there are cultural influences that provide guidelines for what is acceptable sexual behavior, including consent behavior, for both men and women (Gagnon & Simon, 2009, 2005); Sexual scripts often encompass predictive patterns of how sexual encounters progress, how consent is obtained, and how to interpret an individual’s behaviors during a sexual encounter (McCormick, 2010). Sexual scripts tend to be gendered, where men are expected to be sexual aggressors who always want to have sex, whereas women are expected to be the gatekeepers of sexual relations and act resistant to sexual encounters prior to consenting (Gagnon & Simon, 2005). In addition, evidence suggests that exposure to hypermasculine peer norms may lead to higher likelihood of sexual violence toward women (Schwartz & Dekeseredy, 1997). These sexual scripts create an expectation that men and women have opposing goals in terms of sexuality (Gagnon & Simon, 2009), which contradicts the mutual nature of consent and could lead to consent not being respected.
Valid consent is explicit, affirmative, ongoing, mutual, and between competent individuals. Consent is explicit when verbal and nonverbal communication are very clear (Beres, 2007). Consent must also be affirmative, in that a “yes” must be present as opposed to the absence of a “no” (Moore & Reynolds, 2004), and ongoing, meaning that consent must be given throughout the process. Ongoing consent can be done verbally through check-ins throughout the process or through evaluating one’s partner’s behavior for active participation (Beres, 2010). Consent must also be mutual and between competent individuals, meaning that all individuals involved must be of age, mentally fit, and sober (Moore & Reynolds, 2004). If consent is not obtained, a sexual assault has occurred. Studies have shown that men and women express and interpret consent differently (Jozkowski, Peterson, Sanders, Dennis, & Reece, 2014) and that men are more likely to interpret ambiguous sexual situations as consensual (Humphreys, 2007). Thus, it is important to understand how males and females view consent and how they intend to behave in various consent scenarios to work toward preventing sexual assault.
Current Study
It is important for researchers to begin to understand the diversity of sexual scripts and how these scripts contribute to consent behaviors. This knowledge could inform interventions designed to decrease the occurrence of sexual assault on college campuses. Thus, the purpose of this study was to answer the following research questions:
Method
Participants and Procedure
An anonymous survey, based on the White House Task Force to Protect Students from Sexual Assault (2014) survey, was distributed to a random sample of 2,500 students at a Midwestern university. Of the 717 students who completed the survey, 420 identified as female, 287 as male, five as other, and five chose not respond to the question. For the purpose of this study, we excluded individuals who did not respond to this question because gender is central to sexual scripts theory. We also excluded those who did not identify as either male or female because this group was too small to include in the analysis. Most of the participants were White (n = 572), heterosexual (n = 645), and, on average, 21.52 years old. The sample collected is relatively more female than the general student population, but reflected the ethnic distribution of the university.
Measures
Gender was measured by asking participants to choose how they identified their gender. After individuals who did not identify as male or female were removed, this variable was dichotomously coded as male (1) and female (0).
Intentions to respect consent was measured by asking participants a series of four questions revised from the bystander attitudes scale (McMahon, 2010) regarding their likelihood of obtaining consent. Specifically, participants were asked how likely they were to: “Ask for consent to engage in sexual activity with my long-term intimate partner,” “Stop sexual activity when asked to while I am already aroused,” “Continue having sex with a partner after they nonverbally indicate that they are uncomfortable,” and “Decide to have sex with a partner when they are very drunk.” Items were scored on a 5-point Likert-type scale where 1 = very unlikely, 3 = unsure, and 5 = very likely. The last two questions were reverse-coded so that overall higher scores indicated a greater likelihood of respecting consent.
Controls
Past sexual assault victimization was assessed by asking the participants “Have you ever experienced sexual assault as an adolescent or adult prior to coming to college” and “Have you experienced sexual assault since coming to college.” The items were combined into one variable and dichotomously coded as 0 = never experienced sexual assault and 1 = have experienced sexual assault prior to or during college. Participation in previous sexual assault training was assessed by asking the participants “Have you received training from your university in prevention of sexual assault” and coding it as 1 = Yes, 0 = No.
Analysis Plan
A latent profile analysis (LPA; Lanza, Flaherty, & Collins, 2003) was used to determine whether individuals treated gaining consent in sexual relations as uniform or whether there were groups who viewed it differently. LPA is a participant-centered approach that classifies individuals into different groups based on similar characteristics (Lanza et al., 2003). The LPA was conducted via MPLUS (Muthén & Muthén, 1998/2017), utilizing maximum likelihood estimation. After determining variation in the sample, we compared the two profile model to the one profile model and compared each subsequent model until a relatively best fitting model was determined based on theory, whether the model converged, entropy higher than .80 (Clark & Muthén, 2009), smaller log-likelihood, lower adjusted Bayesian information criterion (BIC), a significant Lo–Mendell–Rubin adjusted LRT (likelihood ratio test) test compared with the previous model (Lo, Mendell, & Rubin, 2001), and the bootstrapped LRT test (Arminger, Stein, & Wittenburg, 1999). Once the number of profiles was determined, the mean for each question within each profile was reported. These means were compared with the scale values to contextualize the profiles and determine names. Gender, participation in sexual assault training, and past victimization were then added to the model as predictors of each profile, but sexual assault training and past victimization were removed as the paths were not significant.
Results
The LPA revealed a clear four-profile solution (Table 1 and Figure 1). Individuals in Profile 1 (76.55%), the respecting explicit, affirmative, and ongoing consent profile, strongly agreed that they would ask for consent from their partner before engaging in sexual intercourse (M = 4.77), stop engaging in sex if their partner said to stop (M = 4.93), stop having sex if their partner appeared uncomfortable (M = 5.00), and would not have sex with a person who was drunk (M = 4.29). Individuals in Profile 2 (9.89%), the only respecting explicit consent profile, reported they would ask for consent from their partner before engaging in sexual intercourse (M = 4.42) and would stop engaging in sex if their partner said to stop (M = 4.68). However, they would be very unlikely to stop having sex if their partner appeared uncomfortable (M = 1.00) and were somewhat likely to engage in sex with someone who was drunk (M = 2.49). Similar to Profile 2, individuals in Profile 3 (9.04%), the unsure about nonexplicit consent profile, were very likely to ask for consent from their partner before engaging in sexual activity (M = 4.39) and stop having sex with their partner if he or she said to stop (M = 4.50). However, they were unsure (indicated by a 3) whether they would continue having sex with a person who appeared uncomfortable (M = 2.78) or have sex with a person who was drunk (M = 2.91). Individuals in Profile 4 (4.52%), the does not ask for verbal consent in relationships profile, were similar to those in Profile 1, except they were very unlikely to ask their partner for consent (M = 1.02). They did, however, report they would stop having sex with their partner if they were told to stop (M = 4.62), stop having sex if their partner looked uncomfortable (M = 5.00), and would not engage in sex with a partner who was drunk (M = 4.29). Gender was added as a covariate to determine the likelihood of males and females being in each profile. Female was the reference category in the calculation of odds ratios. Gender was not significantly associated with any profile except Profile 2 where males were significantly more likely to be represented.
Model of Fit Indices for Different Number of Profiles.
Note. BIC = Bayesian information criterion; LMR = Lo–Mendell–Rubin-adjusted LRT; LRT = likelihood ratio test.
Graph of means for each variable in each profile.
Discussion
It is important to note that nearly all participants in this study said they would ask for consent and would stop having sex with a partner who asked them to stop. In addition, the majority of students were in Profile 1, indicating they would respect consent in each of the situations presented. It is possible these individuals might act differently in the moment, as attitudes and actual reported behaviors might differ from one another, but these results indicate students have a clear understanding of the importance of obtaining explicit consent.
Despite this finding, about 10% of the students were likely to have sex with an individual who was very drunk (Profile 2) and 9% were unsure about whether they would have sex with an individual who was very drunk (Profile 3). It is possible that these individuals hold sexual scripts that it is okay to have sexual relations with someone under the influence of alcohol, which can be connected to the normalization of excessive alcohol intake and sexual relations on college campuses (Armstrong, Hamilton, & Sweeney, 2006). This finding also suggests that continued emphasis on the fact that someone who is intoxicated cannot give consent remains important.
In addition to being more likely to engage in sexual activity with a person who is intoxicated, individuals in Profile 2 indicated they would not stop having sex with a partner if he or she appeared to be uncomfortable during the act, and individuals in Profile 3 were unsure what they would do in this situation. Although explicit verbal consent cues are important, ignoring nonverbal cues is not respecting ongoing consent and can be harmful. In addition, sexual scripts theory (Gagnon & Simon, 2005) suggests that women tend to feel pressured to be passive during sex, meaning they might not interject verbally if they are uncomfortable. This suggests that individuals should be encouraged to pay attention to nonverbal cues and follow-up with verbal check-ins to respect ongoing consent. This can also be used to stress that just because someone does not verbally say “no” or “stop,” does not mean they are consenting to sexual activity (Moore & Reynolds, 2004).
Last, we found that 5% of the sample reported they would not ask for verbal consent in long-term relationships (Profile 4). This could be related to previous research that has shown individuals in long-term relationships might utilize nonverbal consent as they learn each other’s signals (Hall, 1998). However, it is also possible that some college students hold sexual scripts that might not fully respect their partner’s ability to say no to sex or they might not believe verbal consent is necessary in relationships. In support of this finding, Katz and Tirone (2010) found that over half of the female college students in their sample had experienced sexual coercion in their relationships. Although we do not know whether this profile is consciously making the decision to not respect consent in their relationships, or whether they are using explicit but nonverbal methods of obtaining consent, it is possible that these individuals are uncertain of the necessity of obtaining consent within romantic relationships. Thus, continued emphasis on consent within relationships is important to stress.
Although one’s knowledge regarding consent might be an important component of how individuals intend to behave, it would be naïve to think participants who indicated their behavioral intentions were to ignore consent in specific scenarios were due to a misunderstanding. Instead, this might be partly due to existing sexual scripts. For example, consent might be viewed as an oppositional experience instead of a collaborative discussion for those who hold the sexual script that men are aggressors and women are gatekeepers of sex (Gagnon & Simon, 2005). In support of this idea, individuals in Profile 2, the only respecting explicit consent profile, were more likely to be men, and might, as part of their sexual script, believe that the only way for a woman to “gatekeep” is through verbal expression.
Limitations and Future Research
Although this study provides evidence for different sexual scripts related to consent, there are limitations. First, the survey had only four questions related to behavioral intentions for obtaining and respecting consent. Future research should assess behavioral intentions in other potential contexts related to consent. Second, the results are subject to volunteer bias as only 29% of the students completed the survey. Third, social desirability bias might be an issue due to the nature of the topic. However, even if the participants answered based on social desirability instead of their own intentions, this still gives insight into how sexual scripts are formed around consent. Fourth, because sexual scripts suggest that males initiate sex, participants might view obtaining consent as the male’s role, resulting in females answering questions differently.
It is also important to note that this study uses a gendered framework and only participants who identified as male or female were included in the analysis; thus, the results of this study cannot be translated to individuals outside this binary. In addition, sexual scripts theory was used to conceptualize sexual relations between females and males and, though gendered expectations around sex may still influence individuals who participate in same-sex sexual encounters, sexual scripts for same-sex relations are different than sexual scripts for different gender relations. Future research would benefit from examining sexual scripts in same-sex relationships or non-cisgender individuals, as this could add complexity to the theory of sexual scripts and how it might contribute to sexual assault. Finally, our sample was a relatively homogeneous sample. It is important to note that while there might be shared messages and experiences for males and females across different cultural groups, the intersectionality of gender with other identities (e.g., race, religion, etc.) might result in different sexual scripts. Future research exploring how these intersecting identities influence sexual scripts and the scripts surrounding consent can continue to inform knowledge about sexual scripts and sexual assault.
Conclusion
Our study found four distinct profiles of college students’ intentions of obtaining and respecting consent during sexual encounters. The majority of the sample respected explicit, affirmative, and ongoing consent. However, approximately one fourth of the sample fell into the other profiles, where students reported being unlikely to or uncertain whether they would respect consent in different situations. Our findings suggest the continued importance of understanding how gender socialization influences one’s attitudes about consent and emphasizing respecting nonverbal cues that a partner is uncomfortable during sex, the importance of consent within romantic relationships, as well as the notion that an individual who is drunk cannot consent to sex.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
