Predictors of Heterosexual Casual Sex Among Young Adults

Abstract

Casual sex is often associated with the young adult stage in the life course. Most recent research on the prevalence, motives, and consequences of heterosexual casual sex has relied on samples of college students, yet students are only a small and advantaged subset of the young adult population. The current study drew on the Toledo Adolescent Relationships Study, which was collected in 2006–2007 and included young adults (ages 18–24 years) whose trajectories reflected a wider spectrum of educational experiences (N = 1,023). We moved beyond prior work by examining both frequency and type of heterosexual casual sex: lifetime vaginal, lifetime oral, and recent vaginal sex. We found that young adults enrolled or who graduated from 4-year educational institutions reported fewer casual sex partners on all three measures compared to participants with-out a high school degree and those with some college experience. Sexual attitudes were key factors mediating the association between educational status and casual sex behavior. These results indicate that programs aimed at encouraging healthy sexual behavior should target individuals who are at risk of not graduating high school because they are at greatest risk of frequent casual sex partners.

Introduction

Most single young adults in the U.S. are sexually active, with 77% reporting that they had sex during the last 12 months (Mosher, Chandra, & Jones, 2005). Some of these sexual experiences have occurred outside of committed, intimate relationships and typically are referred to as casual sex. Although there is a growing body of research that suggests casual sex encounters of heterosexual college students is common (e.g., Grello, Welsh, & Harper, 2006; Regnerus & Uecker, 2011), this may not be reflective of the sexual experiences of heterosexual young adults in general as the majority of individuals aged 18–24 years (59%) are not currently enrolled in higher education (U.S. Census Bureau, 2008). As such, the analyses was based on the Toledo Adolescent Relationships Study (TARS) (N = 1,023) and studied casual sex experiences among young adults with varying educational trajectories. We examined whether higher education status and other predictors were associated with the number of heterosexual casual sex partners using a diverse sample of young adults.

Casual Sex Experiences and Educational Status

Much empirical work on casual sex and the more broadly defined behavior known as “hooking up” (England, Shafer, & Fogarty, 2007) typically has relied on university students enrolled in four-year programs (e.g., Grello et al., 2006; Owen, Rhoades, Stanley, & Fincham, 2010). Focusing on college students has limited the representativeness of such studies because the majority of young adults are not enrolled in 4-year college programs (U.S. Census Bureau, 2008). Research based on a more general sample of young adults is necessary because college samples have likely biased results toward more advantaged young adults’ experiences.

In a recent book that examined how the university experience has changed in recent times, Bogle (2008) argued that the higher education environment was conducive to casual sex. Bogle reported that many students viewed their college years as a time of experimentation (see also Arnett, 2000), including participating in casual sex before entering the “real world ” of responsible adulthood. Bogle concluded that college students” views about sex were influenced, in part, by the ample opportunity to interact with many potential sex partners of similar age and socioeconomic status.

Two longitudinal studies of heterosexual casual sex among American young adults have relied on a broader spectrum of participants, not just those enrolled in college (Bailey, Fleming, Henson, Catalano, & Haggerty, 2008; Eisenberg, Ackard, Resnick, & Neumark-Sztainer, 2009). The Bailey et al. findings were based on a longitudinal sample of 938 first and second graders (in 1993) from the Pacific Northwest who were young adults in 2003–2004. They reported that college attendance was negatively related to reports of ever having casual sex. Bailey et al. stated that young adults” sexual behavior was a continuation of adolescent behavior and that college appeared to have a protective effect on risky sexual behavior during the 6 months following high school graduation. Eisenberg et al. relied on a sample of young adults who were enrolled in Minnesota public high schools during the 1998–1999 school year and did not find a significant relationship between recently experiencing casual sex and education status. Thus, contrary to popular treatments of casual sex, such as Bogle”s study, these findings based on longitudinal data have suggested that such experiences may not be more likely among young people who attended college. However, one limitation of these studies is the measurement of casual sex; neither study assessed number of casual sex partners. The number of casual sex partners rather than whether casual sex ever occurred (e.g., Bailey et al., 2008) or was the most recent sexual experience (e.g., Eisenberg et al., 2009) is more informative in regard to risky behavior because the majority of sexually active adolescents and young adults eventually gain some experience with casual sex (Manning, Longmore, & Giordano, 2005). Thus, the current study contributed beyond prior work by focusing on the number of casual sex experiences reported by young adults.

Liberal Sex Attitudes

Based on empirical evidence and sexual scripts theory, we expected that men compared with women would report, on average, a greater number of casual sex partners. According to sexual scripts theory, society provides the cultural norms and values associated with gender-appropriate sexual behavior (Simon & Gagnon, 1984). Men”s sexual scripts are more often influenced by peers and broader societal messages to adopt values associated with permissive sexual behaviors, such as casual sex (Cuffee, Hallfors, & Waller, 2007; Maccoby, 1998). As a result, men have tended to report more permissive attitudes toward sex than have women (Cuffee et al., 2007; Petersen & Hyde, 2011; Townsend & Wasserman, 2011). Such values and attitudes are often associated with corresponding behavior. Empirical evidence has indicated that male adolescents, for example, were more likely to report casual sex experiences than their female counterparts (Manning et al., 2005). Studies based on college samples have found that this gender gap in casual sex experiences continues into early adulthood, but these gender differences might not be as large as once thought (Conley, 2011; Eisenberg et al., 2009; Grello et al., 2006; Paul, McManus, & Hayes, 2000). Petersen and Hyde”s (2010) meta-analysis of nationally representative samples reported a medium effect size (d = 0.38) for gender differences for the number of casual sex partners with men reporting more partners. In the current study, we expected a gender difference in casual sex behavior with men reporting more partners than women. Further, we anticipated that permissive sexual attitudes would explain, in part, the gender difference in casual sex behavior.

We also extended sexual scripts theory to focus on the education gradient in casual sex behavior. Disadvantaged young adults, such as those who have not completed high school, would likely report more permissive sexual attitudes (Anderson, 1999; National Campaign to Prevent Teen and Unplanned Pregnancy, 2009). Thus, we expected to observe an education gap in permissive or liberal sex attitudes. Furthermore, socioeconomic status is negatively associated with sexual behavior among late adolescents and we expected to find this association among young adults (Meier & Allen, 2008). Furstenberg (2008) has argued that socioeconomically disadvantaged youths, such as those without a high school degree, do not have the structural advantages that foster and sustain more stable intimate relationships. Over time, normative climates among disadvantaged youths have fostered positive attitudes and beliefs about casual sex involvement, especially among men, which increase the likelihood that this occurs within such contexts (Anderson, 1999; Giordano, Longmore, Manning, & Northcutt, 2009). It is likely that these sexual partnering behaviors and attitudes are a consequence of those from disadvantaged backgrounds having less access to more conventional pathways to adulthood, such as enrollment in higher education or stable employment.

Present Study

This study examined the association between education and the number of casual sex partners reported among both men and women and built upon prior work in four key ways. First, school-based sampling frames (e.g., Add Health) potentially have excluded individuals who were not enrolled in school as adolescents. TARS was a educationally diverse sample and being enrolled in school was not a requirement to be included in the sampling frame. Second, it is possible that there are differences between varying higher educational experiences (e.g., 2-year vs. four-year) and casual sex behaviors. The current study examined a detailed measure of higher education status of young adults. Third, prior work that has examined young adults with a broader range of educational backgrounds has measured casual sex in terms of whether it has ever occurred or whether it was the most recent sexual experience (Bailey et al., 2008; Eisenberg et al., 2009). The current study provided a more nuanced understanding of casual sex experience by examining the number and type of casual sex partners.

A final benefit of this study was that we considered structural and attitudinal factors found in prior studies to be related to sex among adolescents or subgroups of young adults. Specifically, employment status may matter because young adults enrolled in school were less likely to be employed (Bureau of Labor Statistics, 2010) and may be in more age-varied environments, presenting fewer opportunities for casual sex. Additionally, adolescents who were raised in single-parent families, step-families, or “other” family forms were more likely to report casual sex experiences compared with adolescents who were raised in two biological parent families (Manning et al., 2005). This association between adolescents” family structure and casual sex behavior was expected to continue into young adulthood. We also anticipated that young adults who lived with parents would report fewer casual sex partners because of greater parental monitoring, which would lead to fewer opportunities to engage in casual sex behavior. Maternal education was included as an indicator of socioeconomic status and thus, we expected casual sex to be negatively related to maternal education. Male adolescents, older youths, and racial minority young adults reported greater involvement with casual sex compared with their counterparts (Bailey et al., 2008; Manning et al., 2005). As such, we anticipated similar findings with respect to the number of casual sex partners reported. Raley, Crissey, and Muller (2007) reported an association between non-romantic sexual behavior and union status among young adults. We expected that those who were in romantic relationships would be less likely to engage in casual sex. Finally, we expected that those who reported casual sex at sexual debut would report more casual sex partners compared to those who did not have casual sex at first intercourse.

In the current study, we had three main hypotheses. First, we expected that young adults who were enrolled in or graduated from 4-year higher education institutions to report the lowest numbers of casual sex partners. Second, we anticipated that females would report lower numbers of casual sex partners compared to their male counterparts. Finally, based on sexual scripts theory, we expected liberal sex attitudes to mediate the relationships between higher education status and casual sex and gender and casual sex.

Method

Participants

To investigate our research questions, the fourth Wave of the TARS was used. TARS is a longitudinal, four Wave dataset originally collected in 2000 based on a random sample derived from the school records of youths in 7th, 9th, and 11th grades in Lucas County, OH. Although school records were used as the sampling frame, school attendance was not a criterion for participation in the original study. Wave II was collected in 2002 and Wave III in 2004. Current analyses were based on participants from Wave IV, which was collected in 2006– 2007. Wave I included a sample of 1,316 youths and, in Wave IV, there were 1,092 valid participants with a retention rate of 82%. Racial minority youth were oversampled. Analysis using census data reported that the TARS sample was similar to the larger Toledo metropolitan area in terms of background socio-demographic characteristics, including general educational levels, median family income, marital status, and racial distribution. Further, the Toledo metropolitan area was similar to the general United States population on the same characteristics (Seffrin, Giordano, Manning, & Longmore, 2012).

We restricted the analytic sample to young adults, aged 18– 24 years (N = 1,068). Participants excluded from the analyses included individuals who were in high school (N = 45). Fortyeight participants or approximately 5% of the sample were missing on several control variables. To account for the missing data, we used multiple imputation with STATA 12. The final analytic sample consisted of 1,023 participants (See Table 1 for descriptive statistics).

Table 1.

Means and percentages for select demographic characteristics for young adults (N = 1,023)

	M	SE
Demographic characteristics
Gender
Female	53%
Male	47%
Age	20.5	0.1
Race
White (comparison)	66%
Black	23%
Other	11%
Higher education status
Less than high school	13%
High school degree not currently enrolled	33%
Community college	13%
Some college not currently enrolled	14%
Bachelor’s (comparison)	27%

Procedure

The questionnaire was primarily self-administered to afford the participants privacy when answering questions about sensitive topics, such as casual sex experiences. Participants were instructed to ask the interviewer for help as needed while responding to questions on a laptop computer, and were taken through instructional screens before beginning the actual self-administered portion of the interview. The TARS was an appropriate dataset to address our research questions for several reasons. First, the TARS data provided detailed measurement on different types of casual sex behavior (oral and vaginal). Second, TARS directly asked how many casual sex partners participants have had in their lifetime and in the last 2 years.

Measures

Lifetime casual sex partners were measured using the following question: “How many different people of the opposite sex have you had vaginal sex with that you weren”t really dating or going out with?” Lifetime casual oral sex partners were measured using the question: “How many different people of the opposite sex have you had oral sex with that you weren”t really dating or going out with?” Recent casual sex partners were measured with the following question: “In the last two years, how many people of the opposite sex have you had vaginal sex with that you weren”t really going out with?” We used these measures for casual sex because prior research has found that casual sex most often occurred with friends or ex-romantic partners(Manning, Longmore, & Giordano, 2006). Based on this, we believed this broad measure better assessed young adult casual sex behavior because it captured more casual sex relationships not just “one nightstands” or sex with a partner that only occurred once.

Education status was measured at Wave IV with the following question: “ How far have you gone in school?” Responses were categorized as: (1) less than high school; (2) high school graduate, but not enrolled in higher education; (3) currently enrolled or graduated from a two-year institution; (4) some college, but not currently enrolled; and (5) currently enrolled or graduated from a four-year institution (comparison group in multivariate analyses).

Gender was coded as a dummy variable where 1 = female and 0 = male. Employment status was classified as currently unemployed (comparison group), employed part-time, or employed full-time.

Liberal sex attitudes were measured at Wave IV and were drawn from the Snyder, Simpson, and Gangestand (1986) scale. This scale measured the extent of agreement with the following six statements: (1)“A person should only have sex with someone they love” (reverse coded); (2) “A person should only have sex if they are married” (reverse coded); (3) “I would have to be committed to a girl/guy in order to have sex with her/him” (reverse coded); (4) “It would be okay to have sex with someone I wasn”t dating;” (5) “I would feel comfortable having sex with someone I was attracted to, but did not know very well;” and (6) “I would be devastated if I got (someone) pregnant at this age” (reverse coded) (alpha = 0.77). The response categories ranged from 1 = Strongly Disagree to 5 = Strongly Agree, so higher scores reflected more liberal attitudes toward sex.

Living with parents was a dichotomous indicator coded “0” if the participant was currently not living with parents and “1” if the participant currently lived in the parental home. Family structure was measured at Wave I and indicated whether the participant lived with two biological parents (comparison group), single parent, step-parent, or “other” family form, such as living with grandparents.

Maternal education was a Wave I variable and was categorized: (1) less than high school; (2) high school; (3) some college; and (4) Bachelor”s degree or more (comparison group). Age was coded as a continuous variable with responses ranging from 18 to 24 years. Race and ethnicity was measured with a series of dummy variables where white was compared to Black, and “Other” racial and ethnic groups. The “Other” category was comprised of 49% Hispanics and 51% other racial groups, such as Asian and multiracial individuals.

Relationship status was composed of three categories: (1) in a marriage or cohabiting union; (2) currently dating; and (3) currently not dating (comparison group).

First sex occurred with a casual sex partner was a dummy variable where 1 = the participant”s first sexual experience occurred with a casual partner and 0 = the participant”s first sexual experience occurred with a romantic partner or the participant did not have sex.

Analytic Strategy

To account for the overdispersion of the dependent variable, negative binomial regression was used (Cameron & Trivedi, 1998). There were two models for each of the three measures of casual sex. The zero-order model included bivariate associations and the full model included the multivariate analysis. Finally, interaction models were tested to determine whether the association between education status and casual sex varied by gender. We determined model fit by comparing whether there was a reduction in the Bayesian Information Criterion (BIC) and the Akaike”s Information Criterion (AIC). Lower BIC and AIC values suggested better model fit. There were no significant interactions between education and gender so the more parsimonious models are shown.¹ In addition, all possible mediations were tested using the analysis suggested by Baron and Kenny (1986).² Negative binomial models can be interpreted more easily by exponentiating the coefficients. The interpretation of the independent variables is that for every unit increase there was an increase in the number of reported casual sex partners by a factor of the exponentiated coefficient. When the coefficients are exponentiated in this way, it is known as the incident rate ratio (IRR). Numbers > 1 suggested positive and numbers <1 suggested negative relationships between the independent and dependent variables.

Results

Descriptive Analyses

The majority 59% of the sample reported some form of casual sex. Specifically, 54% reported having had casual vaginal sex, 44% reported casual oral sex, and 39% reported casual vaginal sex in the past 24 months (results not shown). The mean number of lifetime casual vaginal sex partners reported was 3.1 and lifetime casual oral sex partners was 2.1. Participants reported, on average, 1.4 casual sex partners in the past 24 months (Table 2).

Table 2.

Means for casual sex partners and liberal sex attitudes for young adults (N = 1,023)

	Total	Females	Males
	M (SE)	M (SE)	M (SE)
Lifetime casual sex partners	3.1 (0.2)	2.0 (0.2)	4.3 (0.5)
Lifetime oral casual sex partners	2.1 (0.2)	1.2 (0.1)	3.2 (0.3)
Casual sex partners, past 24 months	1.4 (0.1)	0.8 (0.1)	2.0 (0.3)
Liberal sex attitudesa	16.4 (0.2)	15.0 (0.2)	18.0 (0.2)

^aRange: 6–30

The mean scores for liberal sex attitudes were shown on the bottom row of Table 2. For the entire sample, the mean of 16.4 fell slightly below the midpoint of 18 which indicated slightly conservative sexual attitudes. Female responses (M = 15.0) were below the midpoint and indicated somewhat conservative sexual attitudes and male responses (M = 18) were neutral on sex attitudes.³ Males reported more liberal sexual attitudes than females, but their views fell at the midpoint of the liberal sexual attitude scale.

Bivariate Analyses

At the zero-order level (Table 3), individuals without a high school degree and those with some college experience reported significantly more lifetime casual sex partners compared to individuals who were currently enrolled or graduated from a four-year higher education program. Those with a high school degree or attended community college did not report a difference in the number of life time casual sex partners. As expected, women reported a significantly lower number of lifetime casual sex partners by a factor of 0.47 partners.

Table 3.

Incident rate ratios and standard errors for the zero-order of the negative binomial regressions of number of casual sex partners (N = 1.023)

	Lifetime casual sex partners		Lifetime casual oral sex partners		Casual sex partners during the past 24 months
	Zero-order		Zero-order		Zero-order
	IRR	SE	IRR	SE	IRR	SE
Intercept
Higher education status (bachelor’s comparison)
Less than high school	3.20***	.62	2.02**	.44	2.21***	.49
High school degree not currently enrolled	1.24	.19	0.95	.16	1.27	.22
Community college	1.44	.28	1.08	.24	1.25	.29
Some college not currently enrolled	2.74***	.51	1.85**	.40	1.94**	.42
Gender
Female	0.47***	.05	0.39***	.05	0.40***	.05
Employment status (not employed comparison)
Part-time	0.61**	.09	0.74	.12	0.62**	.10
Full-time	0.78	.11	0.97	.16	0.89	.14
Liberal sex attitudes	1.23***	.01	1.21***	.02	1.26***	.02
Family factors
Live with parents	0.61***	.07	0.73*	.10	0.82	.11
Family structure (two biological comparison)
Single	2.11***	.30	1.21	.20	1.86***	.30
Step	2.39***	.42	2.12***	.42	2.63***	.51
Other	1.95**	.38	1.81**	.40	1.00	.23
Maternal education (bachelor’s comparison)
Less than high school	1.87**	.41	0.98	.24	2.00**	.49
High school	1.27	.21	1.19	.22	1.45*	.27
Some college	1.32	.21	0.94	.17	1.58*	.29
Demographic characteristics
Age	1.26***	.05	1.23***	.05	1.05	.05
Race (White comparison)
Black	2.07***	.29	1.13	.18	1.73**	.27
Other	2.35***	.43	1.91**	.40	1.90**	.40
Relationship status (not in a relationship comparison)
Currently married/cohabiting	1.02	.16	0.73	.13	0.41***	.07
Currently dating	0.78	.11	0.79	.12	0.81	.12
First sex was with a casual partner	2.69***	.34	2.42***	.34	2.16***	.32
Dispersion

^*p<.05;

^**p<.01;

^***p<.001

As Table 3 illustrated, there were other significant control variables at the zero-order. Individuals who were employed part-time compared with those not employed reported significantly fewer lifetime casual sex partners. Liberal sex attitudes were associated with life time casual sex; specifically, for every unit increase in liberal sex attitudes, there was an increase in the reported number of casual sex partners by a factor of approximately 1.2. Participants who currently lived with their parents reported fewer lifetime casual sex partners compared to those who did not reside with parents. Young adults who lived with two biological parents as adolescents reported significantly fewer lifetime casual sex partners. For lifetime casual sex partners, participants whose mothers had less than a high school degree had significantly more partners compared to participants whose mother had a four-year degree. Older participants reported more lifetime casual sex partners, most likely because they had a longer period of exposure to sexual experience. The number of casual sex partners varied by race in that Blacks participants and those classified as “Other”, compared with whites, reported more lifetime vaginal casual sex partners. Finally, young adults who had casual sex at sexual debut reported a higher number of lifetime casual vaginal partners.

Similar to the number of lifetime casual sex partners, participants without a high school degree and those with some college experience reported significantly more lifetime casual oral sex partners compared to individuals who were currently enrolled or graduated from a four-year higher education program. Females had 0.39 fewer lifetime casual oral sex partners compared to males. Liberal sex attitudes, living in a stepfamily or “Other” family form, being older, identifying at “Other” race, and having casual sex at first sex were all positively related to lifetime oral sex partners at the zero-order level. Finally, participants who lived with their parents had significantly fewer casual oral sex partners compared to young adults who did not resided in the parental home.

Lastly, participants without a high school degree or who have some college education had significantly more recent casual sex partners compared to those currently enrolled or with a Bachelor”s degree. As with both lifetime measures of casual sex, females had significantly fewer recent casual sex partners compared to males. Similar to lifetime casual sex partners, participants who were part-time employed had significantly fewer recent partners. Higher liberal sex attitudes, being raised in a single-parents or step-parent household, having a mother without a Bachelor”s degree, not identifying as white, and having casual sex at sexual debut were all positively related to the number of recent casual sex partners. Young adults who were in a co-residential union had fewer recent casual sex partners.

Multivariate Analyses

The influence of higher education on all three measures of casual sex was fully or partially mediated by liberal sex attitudes. First, at the zero-order level, higher education status was significantly associated with lifetime casual sex partners (Table 3). Second, based on a one-way analysis of variance (ANOVA) test, we found a significant relationship between higher education and liberal sex attitudes, F(4, 1016) = 14.23; p <.001; not shown. Specifically, participants with less than a high school degree reported the most liberal sex attitudes (M = 18.75). Participants with some college, but not currently enrolled in school reported the second highest liberal sex attitudes (M = 17.60). Next, those currently enrolled or who had a community college degree reported a mean score of 16.07 and high school graduates reported a mean of 15.93. Finally, participants enrolled in four-year programs reported the most conservative attitudes as reflected by the 15.46 mean score (not shown).Lastly, when liberal sex attitudes and higher education were included in the model, participants who had less than a high school degree and those who had some college, but were not currently enrolled, were no longer statistically different regarding number of lifetime casual oral sex partners and recent casual sex partners compared with individuals who were currently enrolled in four-year institutions. Regarding number of lifetime casual sex partners, liberal sex attitudes partially mediated the effect of higher education status. The significance level for participants who did not have a high school degree or some college experience but were not currently enrolled was reduced, but remained significant in the full model (Table 4).

Table 4.

Incident rate ratios and standard errors for the full model of the negative binomial regressions of number of casual sex partners (N = 1,023)

	Lifetime casual sex partners		Lifetime casual oral sex partners		Casual sex partners during the past 24 months
	Model 1		Model 1		Model 1
	IRR	SE	IRR	SE	IRR	SE
Intercept	0.00***	.00	0.01***	.01	0.03***	.02
Higher education status (bachelor’s comparison)
Less than high school	1.67**	.32	1.11	.26	1.00	.22
High school degree not currently enrolled	1.19	.18	0.76	.14	0.95	.17
Community college	1.39	.25	0.92	.19	1.07	.22
Some college not currently enrolled	1.59**	.27	1.05	.21	1.01	.20
Gender
Female	1.08	.12	0.73*	.09	1.08	.14
Employment status (not employed comparison)
Part-time	0.79	.11	0.93	.14	0.86	.13
Full-time	0.84	.11	0.95	.14	1.06	.15
Liberal sex attitudes	1.19***	.02	1.18***	.02	1.23***	.02
Family factors
Live with parents	0.95	.11	0.94	.13	0.95	.13
Family structure (two biological comparison)
Single	1.45**	.19	0.98	.15	1.35*	.20
Step	1.57**	.24	1.68**	.29	1.91***	.32
Other	1.30	.24	1.55*	.33	0.99	.22
Maternal education (bachelor’s comparison)
Less than high school	0.88	.17	0.45**	.10	1.20	.27
High school	0.73*	.11	0.77	.13	0.90	.15
Some college	0.71*	.10	0.68*	.11	0.91	.15
Demographic characteristics
Age	1.19***	.04	1.13**	.04	0.99	.04
Race (White comparison)
Black	1.59***	.20	0.82	.12	1.61**	.23
Other	1.70**	.27	1.49*	.28	1.26	.23
Relationship status (not in a relationship comparison)
Currently married/cohabiting	0.79	.12	0.57**	.10	0.36***	.06
Currently dating	0.98	.12	1.03	.14	0.93	.12
First sex was with a casual partner	1.86***	.22	1.79***	.24	1.46**	.21
Dispersion	1.70	.12	2.25	.17	1.72	.16

^*p<.05;

^**p<.01;

^***p<.001

Liberal sex attitudes also mediated the association between gender and lifetime casual sex partners and recent casual sex partners, and partially mediated the relationship between gender and number of casual oral sex partners. We found that gender was significantly associated with liberal sex attitudes with men having reported more liberal sex attitudes compared to women, t(1,021) = 10.30; p <.001; not shown. Finally, when liberal sex attitudes were included in the model, women did not report statistically fewer lifetime casual sex partners or fewer recent casual sex partners compared to men, which suggested mediation. Further, gender remained significant in the full model for the number of oral sex partners, but was dramatically reduced, thus partially mediating the association (Table 4).

Finally, some of the control variables remain significant in the full models. Individuals with more liberal sex attitudes, who had casual sex at sexual debut, who were raised in a single-parent or step-family home, who were older, or who self-reported as a racial/ethnic minority still reported significantly more lifetime casual sex partners.

Participants with higher liberal sex attitudes, who were raised in single-parent or step-families, who were older, were classified as “Other” racially, or had casual sex at first sex had significantly more casual oral sex partners. Finally, having high liberal sex attitudes, being raised in a single-parent or step-family, having self-identified as Black or having casual sex at sexual debut was positively associated with the number of recent casual sex partners. Lastly, being a co-residential union was negatively related to recent number of casual sex partners in the full model.

Discussion

The results of the current study showcase the importance of including a diverse sample of young adults when studying casual sex behavior. Specifically, youth who did not have a high school degree or individuals with some college experience had more lifetime and recent casual sex partners. The relationship between higher education status and the number of casual sex partners was mediated or partially mediated by liberal sex attitudes. As expected, females had significantly fewer casual sex partners compared to males. Similar to the higher education status, the relationship between gender and the number of casual sex partners was mediated or partially mediated by liberal sex attitudes.

The current results have highlighted that as adolescents matured into adulthood it was typical to accumulate a greater number of sexual experiences, including, for many, experience with sex outside a traditional dating relationship. These basic findings have underscored the need to extend the reach of prevention and intervention efforts to include young adults as well as adolescents. Such efforts should stress the consistency of condom use and also the varied contexts within which sex occurs. This need is particularly pressing, since increases in the average age at first marriage (U.S. Census Bureau, 2011) have resulted in a longer period of non-marital sexual activity and thus a potentially greater level of exposure to sexual risks.

The current study also highlighted that although prior research on casual sex has tended to focus on college students, the findings have shown that participants with less than a high school degree and those with some college reported significantly more casual sex partners compared to young adults enrolled in four-year college degree programs. Such findings have highlighted the need to develop strategies for reaching non-college youths (whether through family planning agencies or other community settings), those attending community, trade or other technical schools as well as college students.

The results of our study also revealed that programmatic efforts will need to address constellations of attitudes that support involvement in casual sexual relationships, since liberal attitudes were significantly related to variations in reports of lifetime and recent partners. Further, consistent with sexual scripts theories, male participants reported higher levels of casual sex, but it is noteworthy that liberal sex attitudes mediated this gendered relationship for vaginal casual sex partners. Similarly, non-college youth reported more liberal attitudes, which, in turn, mediated the relationship between casual sex and education status for oral casual sex and recent casual sex. Prior research on casual sex has emphasized that the college setting represents an ideal normative climate and physical backdrop (e.g., dormitory life) for casual sexual involvement. Yet, the findings of the current study have suggested that these youth actually reported more conservative attitudes and behaviors, perhaps connecting to their levels of investment in academic and career pursuits. Thus, it is important to develop prevention curricula that raise awareness of disincentives for young adults who are less favorably positioned.

We recognized that participants who reported higher numbers of casual sex partners may also have claimed to have higher liberal sex attitudes. We cannot determine the direction of the causal relationship between attitudes and behaviors but supplemental analysis showed that Wave III liberal sex attitudes was tied to later casual sex behavior at Wave IV. More specifically, we used negative binomial regression and found that Wave III liberal sex attitudes were significantly related to Wave IV number of lifetime casual sex partners (IRR = 1.22; p <.001), number of lifetime oral casual sex partners (IRR = 1.19; p <001), and number of recent casual sex partners (IRR = 1.18; p <.001).

The findings of the current study should be replicated with national samples of adolescents and young adults, although currently available national data sets have not included direct questions about participants” number of casual partners. Further, our study was limited to casual sexual behavior with the opposite sex and within a narrow age range and thus future work should focus on the full array of sexual experiences in young adulthood, including same-sex experiences and other age ranges. It is also important to investigate where and how casual sexual behavior fits with involvement in more serious dating relationships. Some recent treatments of casual sex in the media have described a decline in traditional dating and romance and a concomitant rise of casual “hook-ups,” but research data have suggested that serious relationships are still much in evidence during the young adult period (Lyons, Manning, Longmore, & Giordano, 2010). Nevertheless, contemporary studies have documented a considerable amount of relationship “churning,” that is, breaking up and getting back together, having sex with an “ex,” etc. (see Manning et al.,2006). Thus, more casual encounters may occur during these periods of breakup/instability, which may place more serious partners at risk during periods of reconciliation. Other patterns also warrant additional scrutiny (e.g., sexual involvement over a considerable period of time with a stable, but casual partner; recurrent involvement in casual encounters absent any serious relationships), in order to develop a comprehensive understanding of relationship experiences and risk exposure during this phase of the life course. Even with these limitations, the current study showed that the sexual relationship landscape of young adulthood was complex and the findings moved our understanding of casual sex beyond college-based samples.