Are We Blinded by Desire? Relationship Motivation and Sexual Risk-Taking Intentions during Condom Negotiation

Abstract

Effective condom negotiation skills support better sexual health for both men and women. The current study explored relationship motivation (motivation to establish and maintain long-term romantic relationships), gender, and sexual orientation as factors influencing the condom negotiation process. Participants (177 heterosexual women, 157 heterosexual men, and 106 men who have sex with men) read a vignette describing an encounter with a hypothetical new sexual/romantic partner and responded to embedded items and scales. Stronger relationship motivation predicted increased willingness to have condomless sex among women who perceived greater familiarity with the hypothetical partner. Gender and sexual orientation predicted different preferences for condom insistence strategies. The findings suggest that there are a number of conditions that make it more difficult to recognize risk during a sexual encounter and demonstrate how the process of condom negotiation can be impacted by gender, sexual orientation, and relationship motivation.

Sexually transmitted infections (STIs) are a significant and increasing health concern among young people (Centers for Disease Control and Prevention, 2016; Kerry, Nightingale, Hay, & Oakeshott, 2016; Public Health Agency of Canada, 2015), particularly since untreated infections can lead to reproductive health complications and increase the risk of HIV transmission (Marrazzo et al., 2014; Stamm, 2008). Because condomless sex accounts for the majority of new STIs, including HIV, gaining a better understanding of the reasons for condomless sex is an important step toward reducing the incidence. Most individuals experience a strong, innate need to belong and form social bonds, which impacts and often drives human behavior (Baumeister & Leary, 1995; Buss, 1990). However, this desire to form lasting romantic relationships may influence behaviors (e.g., reducing condom use) that could potentially interfere with other goals, like maintaining sexual health by protecting oneself from unintended pregnancy and STI/HIV transmission. In the current study, we explored whether the motivation to achieve relationship goals (relationship motivation) impacts sexual health decision making and the process of condom negotiation among men and women.

Relationship Motivation and Sexual Health

Men and women enter a sexual encounter with a variety of short- and long-term goals in mind. Short-term goals may include sexual satiation and pleasure, and long-term goals might relate to building a lasting relationship with another person (e.g., Buss, 1998; Zawacki et al., 2009). Through monitoring an ongoing encounter, individuals consciously and unconsciously evaluate whether they will be able to attain their goals, as well as which goals are likely to be met. For instance, a man with the goal to develop a romantic relationship with his date will need to decide whether agreeing to have condomless sex will facilitate his goal to begin a long-term romantic relationship with this partner. His desire for a long-term relationship, however, may compete with his desire to avoid the risks associated with condomless sex (Umphrey & Sherblom, 2007).

For the purposes of this study, we defined relationship motivation as the drive or intention to pursue, establish, and maintain long-term romantic relationships. There is much support for the notion that sexual activity and emotional bonding are deeply linked (e.g., Carter, McNair, Corbin, & Williams, 1999; Diamond & Dickenson, 2012; Filippi et al., 2003). Though relationship formation and maintenance are certainly not the only reasons why individuals choose to have sex (e.g., Simon & Gagnon, 1986; Wood, Desmarais, Burleigh, & Milhausen, 2018), research support for the association between sexual activity and bonding between partners (e.g., Birnbaum & Gillath, 2006) suggests that it is reasonable to view engaging in sexual activity as a viable route toward developing a romantic relationship. As such, concerns about establishing or maintaining a new romantic relationship can increase the perceived importance of managing a potential partner’s perceptions about themselves and the burgeoning relationship (Zawacki et al., 2009). Thus, strong relationship motivation can have a powerful effect on behavior, including whether one chooses to have sex with a partner and how or whether they choose to communicate about topics related to sexual safety (like discussing sexual health history or negotiating condom use; see Holland, Ramazanoglu, Scott, Sharpe, & Thomson, 1992).

Beliefs about condom use, and how a partner may feel about condoms, can impact whether condom use is raised in a sexual encounter. For example, some individuals (even those with favorable condom attitudes) discount the protective benefits of condoms and instead worry that insisting on condoms could interfere with having an enjoyable encounter, or with having sex at all (Downing-Matibag & Geisinger, 2009). Indeed, Buysse and Ickes (1999) noted that the discussion of safer sex topics can create a sense of goal conflict in dating relationships. Many individuals worry that discussing and insisting on condoms will cause their partner to view them as promiscuous or distrustful (Afifi, 1999; Canin, Dolcini, & Adler, 1999; Edwards & Barber, 2010a; Hammer, Fisher, Fitzgerald, & Fisher, 1996; Umphrey & Sherblom, 2007), or they may view such discussions as interfering with intimacy or threatening to the future of their relationships (Strachman & Impett, 2009). Indeed, proposing condom use is seen as less romantic (e.g., Broaddus, Morris, & Bryan, 2010) and, among young gay men, condomless sex can be interpreted as a signal of trust and a way of showing interest in developing a stronger emotional connection (Starks, Pawson, Stephenson, Sullivan, & Parsons, 2018). Similar sentiments are also seen among heterosexual men and women, where desire for intimacy with a partner is associated with decreased condom use (e.g., Gebhardt, Kuyper, & Greunsven, 2003). Thus, concerns about developing romantic intimacy could influence not only whether an individual would choose to begin condom negotiation, but also the strategies they would select and whether or not they persist if their partner is resistant to condom use. These concerns, and their impact on condom negotiation, might be particularly strong among individuals with high relationship motivation.

Condom Negotiation Strategies, Gender, and Sexual Orientation

Unlike other health behaviors (like diet or smoking cessation), condom use requires some amount of dyadic cooperation and negotiation (Metts & Fitzpatrick, 1992). Indeed, communication about safer sex (i.e., condom negotiation) has been identified as a key component for increased condom use (French & Holland, 2011; Noar, Carlyle, & Cole, 2006). Heterosexual men and women are known to invoke a variety of verbal and nonverbal strategies to insist on condom use (De Bros, Campbell, & Peplau, 1994; Noar, Morokoff, & Harlow, 2002) including Seduction (the use of sexual arousal to distract or direct one’s partner to use a condom), deception (using false information to get a partner to use a condom), Withholding Sex (using statements like: “No condom means no sex with me tonight”), or simply Directly Requesting condom use. It should be noted that these strategies include both verbal (e.g., directly requesting condom use) and nonverbal (e.g., increasing a partner’s sexual arousal and then simply applying a condom) behaviors. Past work (Noar et al., 2006) indicates that the strategies most frequently linked with condom use (reported retrospectively) were Withholding Sex, Direct Request, and Seduction. However, the use of these strategies is not consistent across gender, and observed gender differences are not consistent across studies. For example, Holland and French (2012) found no gender difference in the use of the Seduction tactic, whereas De Bros et al. (1994) did. Consideration of potential gender differences in condom negotiation is important because men and women may not experience the same effectiveness for strategies and may favor particular strategies due to their gender socialization and experiences (De Bros et al., 1994; Holland et al., 1992; Noar et al., 2002; Tschann, Flores, de Groat, Deardorff, & Wibbelsman, 2010). For instance, heterosexual women are more likely to rely on verbal forms of condom insistence strategies (Bowleg, Valera, Teti, & Tschann, 2010; Noar et al., 2002), and women may rely upon more direct and assertive verbal strategies than men (i.e., Withholding Sex and Direct Request), since such strategies may help increase their safer sex practices (Holland & French, 2012). However, it is unclear whether differences in strategy preference between men and women are related to other components of the condom negotiation process. For instance, have heterosexual men relied upon less direct and nonverbal strategies for condom insistence because they have a greater expectation that their female partners are interested in condom use?

To our knowledge, there has also been no examination of how men who have sex with men (MSM) may differ in their approach to condom negotiation and insistence compared with heterosexual men and women. This is of particular interest since MSM may not be constrained by gender norms in the same way as heterosexual men (e.g., Malebranche, Gvetadze, Millett, & Sutton, 2012) and are also at greater risk for STIs/HIV (CDC, 2016; Public Health Agency of Canada, 2015).

Rationale for Present Study

Research suggests that relationship motivation can impact sexual health behaviors. For example, individuals with strong relationship motivation may find condom negotiation more difficult, since a desire to seek emotional closeness tends to interfere with engaging in protected sex (Bell, Atkinson, Mosier, Riley, & Brown, 2007). In a study conducted by Zawacki et al. (2009), women who were more strongly motivated by relationship goals were less likely to initiate condom negotiation and, when they did, they selected strategies that would support (rather than undermine) relationship goals (e.g., suggesting to a partner that using condoms would show how much he cares for her – a strategy often termed relationship conceptualizing; e.g., Noar et al., 2002). This fits well with the perspective that women may be more prone to self-sacrificing and investing in nurturing romantic relationships, perhaps even at the cost of their own health (Peasant et al., 2015). However, it is unknown how men with strong relationship motivation approach condom negotiation, and how men and women may change their strategies if their partner is resistant to condom use. Additionally, very little is known about how relationship motivation might impact strategy preference. For instance, would someone high in relationship motivation be more or less likely to select an assertive strategy, like withholding sex?

Condom negotiation is a complex process that is vulnerable to many contextual and individual-level factors present in sexual situations. An exploration of how gender, sexuality, and relationship motivation impact decision making during condom negotiation should significantly improve our understanding of this complex process. Further, as Peasant et al. (2015) discussed, much of the existing work on condom negotiation has several conceptual and methodological limitations, such as utilizing retrospective design, which have made it challenging to explore this construct. In order to overcome such limitations, the current study examined condom negotiation using a vignette to strengthen the ecological validity of the study by presenting participants with a realistic scenario in which they “encounter” an appealing new partner. To our knowledge, the current work is the first to compare the effects of relationship motivation in men and women, and to directly compare sexual decision making and condom negotiation strategies among different gender and sexual orientation groups (i.e., women and men who have sex with men, and men who have sex with women).

The Current Research

The primary objectives of the current study were to examine the differences in condom negotiation among different gender/sexual orientation groups and to assess the impact of relationship motivation on condom negotiation. Our aim was to better understand how each of these factors may contribute to differences in approaches to condom negotiation. Given the previously observed gender differences in condom insistence strategies (e.g., Noar et al., 2002), we tested whether similar differences would be evident in our sample and whether this effect would coincide with a disparity in other aspects related to condom negotiation. Specifically, we hypothesized (H1) that men would perceive their partners as being more interested in condom use, but (H2) that they would be less likely to initiate negotiate condom use compared with women. We also hypothesized that MSM would show a preference for different condom insistence strategies compared with heterosexual women and men (H3), given the differences in socialization and sexual scripts. We also investigated whether relationship motivation would impact an individual’s approach to condom negotiation and whether this might coincide with greater sexual risk-taking intentions. We hypothesized (H4) that participants with stronger relationship motivation would be less likely to select assertive condom insistence strategies (like Withholding Sex), compared with those with low relationship motivation. Additionally, based on the previous findings of Zawacki et al. (2009), which suggested that women motivated by relationship goals may be less successful at condom negotiation, we hypothesized (H5) that participants with stronger relationship motivation would also be more willing than those with low relationship motivation to engage in hypothetical condomless sex, regardless of their selection of insistence strategies.

Method

Participants

Eligible participants were between 18 and 25 years of age and indicated that they were not currently involved in a long-term monogamous relationship and that they had engaged in consensual penetrative intercourse at least once. Because sexual safety practices differ greatly for women who have sex with women, only cis-gendered (i.e., individuals whose current gender identity matches the sex they were assigned at birth) female participants who self-identified as being attracted to men (WSM) were eligible to participate, whereas cis-gendered male participants were eligible if they identified as being attracted to women (MSW) or men (MSM) (see Table 1 for demographics).¹

Table 1. Demographics summary

		MTurk Sample	Pool Sample
Variable	Response	n/% of N	n/% of N
Relationship Status	Single	260/85%	85/64%
	New/Casual	46/15%	24/18%
	Relationship less than 1 year	0	22/17%
	Relationship 1 year or more	0	2/2%
Sexual Relationship Status	Not having sex	130/43%	49/37%
	Having sex, but no exclusive partner	139/46%	43/32%
	Exclusive relationship with outside partners	8/3%	2/2%
	Exclusive relationship with no outside partners	29/10%	29/10%
Country of Residence	UK	2/1%	0
	USA	303/99%	0
	Canada	0	133/100%
Ethnicity	White	219/72%	109/82%
	Black	28/9%	6/5%
	Latin	14/5%	2/2%
	Multi-ethnic	17/6%	4/3%
	Other	28/9%	12/9%
Condom Use Past 3 Months	Sometimes or less frequently	103/34%	63/48%
	Often or more frequently	201/66%	70/53%

Participants (N = 440) were recruited using two different recruitment strategies in an effort to achieve a broader sample. Using Amazon’s Mechanical Turk system (Mturk) 307 participants (101 WSM, 104 MSW, and 102 MSM) were recruited. Additionally, 133 participants (76 WSM, 53 MSW, and 4 MSM) were recruited using the participant pool system at a southern Ontario university in Canada. Mturk participants received a small financial incentive (75¢) and university pool participants received a 0.5 bonus credit that could be applied to any eligible course. The two samples were combined, as preliminary analyses revealed response patterns to target items for subsequent analyses were not significantly different.

Materials

Hypothetical Scenario

Participants were invited to read and project themselves into a vignette describing a romantic encounter with a hypothetical new sexual partner (see online supplementary file). The scenario began with meeting an appealing new partner at a party and led to condom negotiation during a sexual encounter. Because the hypothetical partner’s sexual health history was unknown, condomless sex should objectively be considered a risky decision (Comer & Nemeroff, 2000). Participants rated their attitudes and likelihood of choosing particular courses of action in response to items embedded in the scenario. For instance, they were asked to indicate their own and to estimate their partner’s interest in condom use, how willing they would be to introduce the topic of condom use, and what strategy they would use to insist on condom use (if any). During the scenario, it was made clear that a condom may not be available and the hypothetical partner attempts to convince the participant to have condomless sex. After this, participants are asked questions about how risky they think condomless sex with the hypothetical partner would be and how likely they would be to agree to have condomless sex. All embedded items were scored using a 10-point or 100-point Likert type scale (e.g., 0 = “not at all likely”; 10 = “extremely likely”).

When asked to rate their immersion, participants reported finding it easy to project themselves into the scenario (M = 8.59 on a 10-point scale, SD = 1.73). No significant difference was found between the gender/sexuality groups in this rating (p > .05).

Condom Influence Scales

At two different points in the scenario, participants were asked to select a strategy (from a list of options) that they would invoke to convince the hypothetical partner to use a condom during their sexual encounter. Different items were presented at Time 1 and at Time 2 for novelty. The options were derived from the Noar et al. (2002) methods of condom use influence and represented the following dimensions: No Strategy (i.e., agreeing to condomless sex), Withholding Sex, Relationship Conceptualizing, Direct Request, Seduction, Deception, and providing Risk (STI) Information. One item representing each of these dimensions was selected from the highest loading items for each dimension (see Noar et al., 2002) to present seven possible response options at Time 1 and Time 2, respectively (see online supplementary file). These options were embedded in the scenario: Time 1 was presented near the start of the sexual encounter phase of the scenario and Time 2 was presented after it was established that the hypothetical partner is resistant to using a condom.

Relationship Motivation Scale

The Relationship Motivation Scale (RMS), adapted from Sanderson and Cantor (1995) and Kindelberger and Tsao (2014), is scored on a 7-point Likert scale. It consists of 15 items examining motivation to form and maintain long-term romantic/dating relationships (e.g., “In my dating relationships, I typically try to spend a substantial amount of time with my girl/boyfriend(s)”), as well as elements of anti-motivation (e.g., “All things considered, it is better to be alone”), which were reverse coded. Higher scores indicate greater relationship motivation. During pilot testing (N = 124; 68 women), the RMS showed good reliability (alpha = .87). Average inter-item correlation was low (.34), with a range from .02 to .64, suggesting the items measure unique aspects of the construct of relationship motivation. In the current study, the RMS was also found to have good reliability (alpha = .85).

Procedure

Participants who clicked on the invitation link viewed a brief introduction screen and consent form. WSM and MSM participants then read a scenario depicting a romantic encounter with a hypothetical male partner; the MSW scenario depicted a hypothetical female partner. After completing the scenario and embedded questions, a subset of participants (n = 133) were asked to recall and rate how familiar the hypothetical partner would have felt at the start of the condom negotiation portion of the scenario. This was added because familiarity can be built quite quickly (Swann, Silvera, & Proske, 1995) and we wanted to determine whether this had an impact on participants’ decision making in the scenario (e.g., Sparling & Cramer, 2015). All participants then completed the RMS and provided demographic information.

Data Analysis

Three predictor variables were simultaneously entered into multiple linear regression analyses: Relationship Motivation Scale Score (RMSS) and two of the gender/sexuality groups: MSM and MSW (WSM was used as a base category) (see Table 2 for results summaries).² The two outcome variables related to our hypotheses were responses to the items assessing participants’ estimate of the hypothetical partner’s interest in condom use (H2), and participants’ intentions to willingly engage in condomless sex with the hypothetical partner (H5). Two additional outcome variables were assessed in exploratory analyses: (a) participants’ perception of how risky condomless sex with the hypothetical partner would be and (b) their own interest in condom use. ANOVAs were conducted to examine gender/sexual orientation group differences between MSW, MSM, and WSM on the outcome variables noted above. In addition, t-tests were conducted to examine the role of perceived partner familiarity (in the subsample of participants who received this item) on these.

Table 2. Linear regression results

	Partner Interest in Condom Use		Condomless Sex Willingness		Own Interest in Condom Use		Risk Perception
Variables	B (SE)	β (CI: 95%)	B (SE)	β (CI: 95%)	B (SE)	β (CI: 95%)	B (SE)	β (CI: 95%)
MSM	19.56 (3.67)	.273*** (12.36, 26.77)	1.24 (.36)	.167** (.53, 1.95)	−7.23 (3.54)	−.102* (−14.19, −.27)	−.29 (.28)	−.052 (−.83, .26)
MSW (vs. WSM)	10.80 (3.66)	.169** (4.75, 17.94)	2.69 (.32)	.406*** (2.06, 3.32)	−20.63 (3.16)	−.327*** (−26.84, −14.42)	−1.25 (.25)	−.255*** (−1.74, −.76)
Relationship Motivation	1.99 (1.80)	.051 (−1.56, 5.53)	−.57 (.17)	−.142** (−.92, −.23)	4.25 (1.74)	.111* (.83, 7.68)	.45 (.14)	.152** (.18, .72)
Adj. R^2	0.059		0.154		0.097		0.075
F Statistic	10.15***		27.53***		16.76***		12.80***
d	0.063		0.182		0.108		0.081

*p < .05, **p < .01, ***p < .001

Because responses to the condom insistence strategies at Times 1 and 2 were categorical variables with a single response for each participant, these were analyzed using chi-square tests of independence (H1 and H4). McNemar’s test was used across gender/sexuality groups to detect statistically significant differences in participants’ changes in strategy use between Time 1 and Time 2. An independent one-way ANOVA was conducted to examine gender/sexuality group differences in participants’ willingness to initiate condom negotiation (H3). An independent 2 by 3 ANCOVA was conducted in order to investigate whether participants with higher RMSS would be more willing to engage in condomless sex with the hypothetical partner (H5).³ Missingness in the analytic sample was assessed and was determined to be very low (<1% of values) and non-problematic (missing at random). Based on this assessment, multiple imputation was not used for this dataset.

Results

Hypothesis Testing

Hypothesis 1:

Perceived Interest in Condom Use

No significant difference in perceived partner interest in condom use was found based on high versus low familiarity rating (p > .05). Male gender (MSW or MSM) was associated with increased perceived partner interest in condom use (see Table 2). RMSS did not emerge as a significant predictor (see Table 2). A significant difference was also found in participants’ estimation of their hypothetical partner’s interest in using a condom by gender/sexuality group, using an independent one-way ANOVA: F (2, 433) = 15.93, p < .001, d = .26. A post hoc Tukey test indicated that female participants (M = 51.25, SD = 30.69) perceived the hypothetical partner as having significantly lower condom use interest than did either MSW (M = 62.46, SD = 29.78, p < .01, 95% CI: [−18.54, −2.81]) or MSM (M = 71.34, SD = 26.88, p < .001, 95% CI: [−27.68, −10.99]). MSM also perceived more partner interest in condom use than MSW (p < .05, 95% CI: [.16, 17.16]). (See exploratory section for additional analyses comparing participant interest in condom use with perceived partner interest in condom use.)

Hypothesis 2:

Participants’ Reported Likelihood of Initiating Condom Negotiation

Gender/sexuality group differences in sexual health decision making were also examined using responses to an item asking participants to estimate how likely they would be to bring up using a condom during the hypothetical scenario. An independent one-way ANOVA indicated a significant difference based on gender/sexuality for participants’ willingness to bring up condom use, Welch F (2, 257.61) = 7.76, p < .01, d = .19); the nonparametric Kruskal–Wallis test (in response to Levene’s detection of homogeneity of variance; F (2, 436) = 39.5, p < .001) confirmed this result: H (2) = 20.22, p < .001. A set of three nonparametric Mann–Whitney U tests (comparing WSM with MSW, MSW with MSM, and women with MSM) with a Bonferroni correction (alpha set to .016) showed that MSW (M = 71.22, SD = 33.45) were significantly less likely to bring up condom use than WSM (M = 84.22, SD = 25.76) (U = 10,020.00, N₁ = 157, N₂ = 176, p < .001). There was a marginally significant difference found between MSW and MSM (M = 79.06, SD = 27.96) (U = 7,087.00, N₁ = 157, N₂ = 106, p < .05) and between WSM and MSM: U = 8,140.50, N₁ = 106, N₂ = 176, p = .057.

Hypothesis 3:

Condom Negotiation Strategies by Gender/Sexual Orientation

Pearson’s χ² analysis of Time 1 responses indicated a significant effect of gender/sexuality group: χ² (12, N = 438) = 117.89, p < .001, φ = .352 (accounting for 12.4% of the variance). Specifically, WSM were significantly less likely than either MSM or MSW to choose No Strategy (i.e., were less likely to choose to have condomless sex). Additionally, women were less likely than either MSW or MSM to choose the Seduction method (“I would just keep fooling around and then just put a condom on him when it’s time”). However, WSM were significantly more likely than MSW or MSM to select methods such as Withholding Sex (“I would make it clear that we’re not having sex without a condom”), Direct Request, and Deception (“I would make up a reason why I want to use a condom tonight, even though my real reason is to protect myself against diseases”). MSM were significantly more likely than MSW or WSM to select the Relationship Conceptualization method (“I would tell Chris that it would mean a lot and show me how much he cares, if we were to use a condom with me tonight”). Finally, MSW were significantly more likely than MSM or WSM to select No Strategy (“I would be comfortable having condomless sex tonight”) and the Seduction method. MSW were also significantly less likely than MSM or WSM to select the Withholding Sex method or the Direct Request method (see Table 3).

Table 3. McNemar test results for condom insistence strategy change between Time 1 and Time 2

	Agreeing to Condomless Sex			More Assertive Strategy			Seductive Strategy			Deceptive Strategy			Relationship Strategy			Risk Info Strategy
	Groups			Groups			Groups			Groups			Groups			Groups
Discordant Pairs	MSW	MSM	WSM	MSW	MSM	WSM	MSW	MSM	WSM	MSW	MSM	WSM	MSW	MSM	WSM	MSW	MSM	WSM
Present at Time 1 & Absent at Time 2	1	2	0	21	21	40*	41***	26*	11	5	0	11	3	7	4	3	3	6
Absent at Time 1 & Present at Time 2	15**	2	3	22	22	21	13	12	20	11	10**	15	6	2	4	4	9	6

*** p < .001, **p < .01, * p < .05

Later in the scenario (Time 2), participants were again asked to select a strategy, this time to convince their resistant hypothetical partner to use a condom. Pearson’s χ² indicated a significant effect of gender/sexuality group at Time 2: χ² (14, N = 439) = 79.70, p < .001, φ = .294 (accounting for 8.6% of the variance). At Time 2, WSM were less likely than MSM or MSW to select No Strategy (agreeing to condomless sex) or the Seduction strategy and WSM were again more likely to select Withholding Sex as a strategy. MSM were significantly more likely to select the Deceptive strategy (“I would tell Chris that I always have sex with condoms, even though sometimes I don’t”). Finally, MSW were less likely to select Withholding Sex and were more likely to select No Strategy or to select the Seduction strategy (See Table 3).

Changes in strategy use between Time 1 and Time 2 were observed for several of the strategies, although not consistently across the participant groups. There were significantly more MSW who changed their strategy to agreeing to condomless sex between Time 1 and Time 2 (p < .01), compared with those who changed away from this option. This trend was not found among MSM (p = 1.00), nor among WSM (p = .25). There were significantly more WSM who changed their strategy from a more assertive strategy (withholding sex or direct request) to something else between Time 1 and Time 2 (p < .05), compared with those who changed to use this strategy between Time 1 and Time 2. This trend was not found among MSM (p = 1.00), nor among MSW (p = 1.00). There were significantly more MSW who changed their strategy from a seductive strategy to something else between Time 1 and Time 2 (p < .001), compared with those who changed to use this strategy. This trend was also found among MSM (p < .05), but not among WSM (p = .15). There were significantly more MSM who changed their strategy to a deceptive strategy between Time 1 and Time 2 (p < .01), compared with those who changed away from this strategy. This trend was not found among MSW (p = .21), nor among WSM (p = .56). No significant differences were found in changes related to the Relationship Conceptualizing or Risk Information strategies: MSW (p = .51, p = 1.00), MSM (p = .18, p = .15), WSM (p = 1.00, p = 1.00). See Table 4 for cross-tab results of strategies changes.

Hypothesis 4:

Relationship Motivation and Condom Negotiation

RMSS did not differ by gender/sexuality identity (p > .05). To allow for comparison between those who scored higher versus lower on our measure of relationship motivation on subsequent analyses, RMSS was dichotomized. Scores of 5.01 and higher were categorized as High Relationship Motivation (HRM; n = 304) and scores of 4.99 and lower were categorized as Low Relationship Motivation (LRM; n = 122). Participants whose score was exactly 5 were excluded (n = 13), since this was the midpoint of the scale.

Pearson’s χ² indicated a significant effect of RMSS at Time 1: χ² (6, N = 438) = 15.68, p < .01, φ = .189 (accounting for 3.57% of the variance at Time 1). Specifically, it was found that LRM participants were significantly more likely (than HRM) to select No Strategy (i.e., were more willing to have condomless sex). Pearson’s χ² indicated a marginally significant effect of RMS at Time 2: χ² (7, N = 439) = 13.80, p = .055, φ = .177 (accounting for 3.13% of the variance at Time 2). Specifically, it was found that HRM participants were somewhat less likely (than LRM) to select Deception as a condom negotiation strategy and were more likely to select the Direct Request strategy (see Table 5).

Table 4. Summary of condom insistence strategies: Gender/sexuality identity

	Time 1			Time 2
Response	MSW n (%)	MSM n (%)	Women n (%)	MSW n (%)	MSM n (%)	Women n (%)
None - Would Have Condomless Sex	26 (17)	13 (12)	3 (2)	40 (26)	13 (12)	6 (3)
Withholding Sex	18 (12)	22 (21)	70 (40)	20 (27)	29 (27)	81 (46)
Direct Request	31 (20)	23 (22)	59 (34)	28 (18)	17 (16)	29 (17)
Seduction	69 (44)	34 (32)	15 (9)	42 (27)	20 (19)	24 (14)
Relationship Conceptualizing	4 (3)	8 (8)	4 (2)	7 (5)	3 (3)	4 (2)
Risk (STI) Information	3 (2)	3 (3)	9 (5)	4 (3)	9 (9)	9 (5)
Deception	6 (4)	3 (3)	15 (9)	13 (8)	13 (12)	19 (11)

Hypothesis 5:

Willingness to Engage in Condomless Sex

An independent t-test showed that participants were significantly more willing to engage in condomless sex with the hypothetical partner if they rated the hypothetical partner as More Familiar (M = 6.1, SD = 2.9) than Less Familiar (M = 4.1, SD = 3.0): t (117) = 3.62, p < .001, d = .68, 95% CI: [−3.02, −.88].

Table 5. Summary of condom insistence strategies: Relationship motivation

	Time 1		Time 2
Response	Low RM n (%)	High RM n (%)	Low RM n (%)	High RM n (%)
None - Would Have Condomless Sex	20 (16)	22 (7)	21 (17)	38 (12)
Withholding Sex	31 (25)	79 (25)	37 (30)	93 (29)
Direct Request	25 (21)	88 (28)	13 (11)	61 (19)
Seduction	36 (30)	82 (26)	28 (23)	58 (18)
Relationship Conceptualizing	2 (2)	14 (4)	5 (4)	9 (3)
Risk (STI) Information	1 (1)	14 (4)	9 (7)	13 (4)
Deception	7 (6)	17 (5)	6 (5)	39 (12)

Lower relationship motivation and male gender (MSW and MSM) were associated with an increased willingness to engage in condomless sex (See Table 2). An independent 2 (Relationship Motivation: LRM vs. HRM) by 3 (gender/sexuality group: MSM vs. MSW vs. WSM) ANCOVA was conducted. Whether or not participants selected an assertive condom insistence strategy (i.e.: Withholding Sex or Direct Request) at Time 1 or Time 2 were entered as covariates into the ANOVA model. As these options have been defined as more assertive strategies (see Noar et al., 2002; Peasant et al., 2017) they can be seen as an indirect measure of condom use self-efficacy and assertiveness. These were significantly related to participants’ willingness to engage in hypothetical condomless sex [Time 1: F (1, 431) = 6.87, p < .01; Time 2: F (1, 431) = 51.36, p < .001]; participants who selected a more assertive strategy were less likely to indicate a willingness to have condomless sex [Time 1: assertive strategy M = 3.58, SD = 2.70; non-assertive M = 5.58, SD = 3.33; Time 2: assertive strategy M = 3.14, SD = 2.55; non-assertive M = 5.80, SD = 3.16]. Controlling for these variables, a significant main effect of gender/sexuality group emerged [F (2, 431) = 16.85, p < .001, d = .28] as well as a marginally significant main effect of RMSS [F (1, 431) = 3.28, p = .07, d = .09].

A post hoc Games–Howell analysis indicated that MSW (M = 6.00, SD = 3.22) reported being significantly more likely to engage in condomless sex with the hypothetical partner than either WSM (M = 3.30, SD = 2.60, p < .001, 95% CI: [1.42, 2.87]) or MSM (M = 4.60, SD = 3.1, p < .01, 95% CI: [.38, 1.85]). Additionally, MSM reported being significantly more likely to engage in condomless sex than WSM (p < .01, 95% CI: [.28, 1.77]). Furthermore, LRM participants were marginally significantly more willing to engage in condomless sex with the hypothetical partner than HRM participants (p = .07, 95% CI: [−.05, 1.11]).

There was a significant interaction between gender/sexuality and RMSS [F (2, 431) = 4.07, p < .05, d = .14]. The dataset was split by gender/sexuality and an independent t-test was conducted comparing HRM and LRM response patterns. Among MSW and MSM, LRM participants (MSW: M = 6.76, SD = 2.82; MSM: M = 5.42, SD = 3.0) were significantly more likely to show an intention to engage in condomless sex in the hypothetical scenario (MSW: t (87.3) = 1.99, p = .05, d = .34, 95% CI: [.002, 2.15]; MSM: t (101) = 2.06, p < .05, d = .43, 95% CI: [.05, 2.58]) than HRM (MSW: M = 5.72, SD = 3.33; MSM: M = 4.10, SD = 3.13). No significant difference (p > .05, 95% CI: [−1.22, .59]) was found in WSM; however, the means showed a trend toward a slightly higher interest in having condomless sex among HRM WSM.

We tested whether perhaps partner familiarity might play a role, given the results of the familiarity analysis noted above. Thus, an exploratory analysis was conducted using the subsample who rated the hypothetical partner as being more familiar: HRM WSM (M = 6.14, SD = 2.2) were significantly more willing to engage in condomless sex with the hypothetical partner than LRM women (M = 2.6, SD = 3.1); t (17) = 2.79, p < .01, d = 1.32, 95% CI: [−6.22, −.87]. No such difference was found in MSW or MSM in this subsample.

Additional Exploratory Analyses

Participant Reported Interest in Condom Use

Participants were much more interested in using a condom if they rated the hypothetical partner as feeling Less Familiar (M = 85.75, SD = 27.54) than More Familiar (M = 70.90, SD = 34.66): t (93.26) = 2.51, p < .05 (p = .014), 95% CI: [4.32, 28.09], d = .51; this result was confirmed by a nonparametric Mann–Whitney U test to account for Levene’s detection of variance homogeneity (F (1, 432) = 7.92, p < .01). Higher RMS was associated with an increased reported interest in condom use. Male gender (MSW or MSM) was associated with decreased stated interest in condom use (see Table 2).

An independent one-way ANOVA indicated a significant difference based on gender/sexuality for participants’ own interest in condom use, Welch F (2, 240.67) = 19.99, p < .001 (Levene (2,431) = 41.34, p < .001, d = .034). A post hoc Games–Howell test indicated that MSW (M = 69.45, SD = 35.88) were less interested in using a condom for sex with the hypothetical partner than either WSM (M = 89.98, SD = 20.33, p < .001, 95% CI: [−28.75, −13.06]) or MSM (M = 83.32, SD = 24.78, p < .01, 95% CI: [−22.43, −3.94]); this result was confirmed using the nonparametric Kruskal–Wallis test, H (2) = 41.34, p < .001.

Correlation between Self-Interest in Condom Use and Perceived Partner Interest

We next examined the degree of correlation between participants’ interest in condom use and their estimation of the hypothetical partner’s interest. In MSW, a significant moderate correlation was found: higher personal interest in condom use was associated with higher perceived partner interest in condom use: r_p (157) = .56, p < .001. In MSM, a significant strong correlation was found: r_p (106) = .74, p < .001. In WSM, a significant but weak correlation was found: r_p (176) = .34, p < .001. Using the Fisher r-to-z transformation tool (Lowry, 2001), the correlation coefficient for WSM was significantly smaller than for either MSW (z = 2.52, p < .01) or MSM (z = 4.79, p < .001), and the correlation coefficient for MSM was significantly larger than for MSW (z = 2.50, p < .01).

Risk Perception

Risk perception was assessed using ratings of how risky participants felt it would be to have condomless sex with the hypothetical partner. Stronger relationship motivation was associated with an increased perception of risk. Being MSW (but not MSM) was associated with decreased risk perception (see Table 2). An independent one-way ANOVA indicated that members of different gender/sexuality groups showed significantly different levels of risk perception: Welch F (2, 247.60) = 12.47, p < .001, (Levene (2, 434) = 11.12, p < .001), d = .24; this result was confirmed by a Kruskal-Wallis test: H (2) = 20.67, p < .001. A set of two Mann-Whitney U tests (comparing WSM with MSW and MSM with MSW) were conducted (alpha set to .025): MSW (M = 7.04, SD = 2.68) perceived significantly less risk associated with engaging in condomless sex than WSM (M = 8.32, SD = 1.87) (U = 9,861.50, N₁ = 156, N₂ = 175, p < .001, 95%CI: [−1.89, −.68]) as well as MSM (M = 7.98, SD = 1.87) (U = 6,526.00, N₁ = 156, N₂ = 106, p < .01, 95%CI: [−1.67, −.21]).

We speculated whether participants’ perception of the risk associated with having condomless sex with the hypothetical partner might mediate their estimation of their willingness/intention to have condomless sex in the scenario. A series of mediation analyses were conducted using the Process module in SPSS (Hayes, 2012) in order to explore the potential mediating effect of risk perception on the association between relationship motivation and intention to have condomless sex and between gender/sexuality and intention to have condomless sex. Evidence of complete mediation was found for RMSS, whereby the association between relationship motivation and intention to have condomless sex was no longer significant when risk perception was included in the model (β = −.653, SE = .175, 95% CI: [−.992, −.300]). No mediation effect was detected for MSM, but there was evidence of partial mediation for WSM: the association between gender (among WSM and MSW) and intention to have condomless sex in the scenario was partially mediated by risk perception (β = −1.791, SE = .242, 95% CI: [−2.255, −1.311]). The data suggest that HRM participants and female participants each tended to perceive greater risk associated with condomless sex in the scenario, which, in turn, led to a decreased willingness to engage in condomless sex in the scenario.

Risk perception was also found to fully mediate the effect of gender among MSW and WSM on interest in condom use (β = 12.01, SE = 2.03, 95% CI: [8.24, 16.09]) and partially mediated the likelihood of selecting a more assertive strategy (i.e., Withholding Sex or Direct Request) at Time 1 (β = .344, SE = .087, 95% CI: [.199, .538]) and at Time 2 (β = .574, SE = .121, 95% CI: [.362, .834]). This effect was not replicated in the MSM sample. This result (see Table 3) suggests that women tended to perceive greater risk associated with condomless sex, which was, in turn, associated with an increased desire to use condoms and a greater likelihood of relying on more assertive condom insistence strategies.

Discussion

The purpose of the current work was to examine condom negotiation in a sample of WSM, MSW, and MSM. Participants were presented with a tailored hypothetical scenario and responded to embedded items, as well as a measure of relationship motivation. The results indicated that, while both groups of men were more willing to engage in condomless sex than the female participants, men who scored higher in relationship motivation (compared with those who scored lower) were less willing to do so. Further, women who scored higher in relationship motivation were more willing to engage in condomless sex in the hypothetical scenario (compared with women who scored lower), but only if they also viewed the hypothetical partner as more familiar. This finding suggests that gender/sexuality and relationship motivation have significant implications for sexual health decision making in the condom negotiation process.

Gender/Sexuality Differences in Condom Negotiation

A significant difference was found in perceived and stated condom interest between male and female participants (H2): both male samples perceived higher partner interest in condom use than WSM (which is in agreement with the findings of Edwards and Barber (2010b) and Tschann et al. (2010). We also found a significant difference between male and female participants in their willingness to initiate condom negotiation (H3): both groups of men reported a lower likelihood of suggesting condom use than WSM. Interestingly, MSW were somewhat less likely to suggest condom use than MSM. MSW also demonstrated a significantly lower stated interest in condom use during the hypothetical scenario than either WSM or MSM. This is seen not only in MSW responses to the condom interest item, but also in the condom negotiation strategies preferred by MSW. At Times 1 and 2, MSW were most likely to select no strategy (agreeing to condomless sex).

This is the first study, to our knowledge, to compare condom use intentions among WSM, MSW, and MSM and the correlations between personal interest in condom use and perceived partner interest illustrates the differences between these three groups. Among all samples, a significant correlation was found – suggesting that participants tend to assume their partners have similar attitudes as themselves to safer sex. However, the correlation was significantly stronger in MSM and significantly weaker in WSM. This suggests that MSM may have much more confidence that partners’ safer-sex intentions will match their own intentions than do WSM (regardless of whether such confidence is warranted).

As predicted, MSW, MSM and WSM showed group differences in their approach to condom negotiation (H1). This is the first time, to our knowledge, that condom insistence strategies were compared among WSM, MSW, and MSM. We found that, besides being more prepared to agree to condomless sex, when MSW did select a condom negotiation strategy, their strategies were most likely to involve a Direct Request or a Seduction strategy. This finding fits with the assertion of Noar et al. (2002) that MSW have an advantage in condom negotiation – which they seemed to apply in their selection of nonverbal condom negotiation strategies in the current study (see also Allen, Emmers-Sommer, & Crowell, 2002; De Bros et al., 1994). As the wearer of the condom, MSW may find that the Seduction method for condom insistence is both effective and less confrontational. When applied by MSW in this context, this strategy may be better framed as a more passive and normative strategy than a purposely “seductive” one. Simply donning a condom at the appropriate moment (however this moment is determined) would likely not be done in order to manipulate the situation or sexual partner(s), but rather to introduce the prophylactic innocuously, without disrupting the mood of the encounter. It is also interesting to note that more MSW changed from (than changed to) using a seductive strategy at Time 2 and more changed to (than changed from) agreeing to condomless sex. This may be reflective of MSWs’ preference for passive condom use strategies, as agreeing to the hypothetical partner’s request for condomless sex would be the most passive option of all.

In contrast, WSM were most likely to use more assertive strategies, like Withholding Sex. It is possible that this response pattern is due to some degree of socially desirable responding, particularly among women who experience greater societal pressure to enforce condom use. However, the current results do fit with previous work showing that WSM tend to take a more active role in condom negotiation than MSW (Carter et al., 1999) and that women are more likely to use more assertive strategies with casual partners (De Bros et al., 1994; Holland & French, 2012). Interestingly, WSM endorsed a wider variety of strategies (than MSM or MSW) at Time 1 than at Time 2. However, the results of the McNemar’s test indicated that more women were switching to a strategy other than the most assertive options (Withholding Sex or Direct Request) than were changing to select these strategies between Times 1 and 2. This is concerning because work by Peasant et al. (2017) has shown that condom use was more likely when women applied a Withholding Sex condom insistence strategy. Future research should explore how women decide whether to change strategies and which strategies are more likely to be selected when a woman perceives an assertive strategy as unsuccessful; these questions were beyond the scope of the current analysis.

Women are faced with gender-based power differentials in many arenas (see Horton & Dworkin, 2013), particularly during the negotiation of sexual safety practices. Overall, the WSM in this sample did select more assertive strategies, and it may be that this was a response to previous experience with unwanted condomless sex or condom coercion (see Teitelman, Tennille, Bohinski, Jemmott, & Jemmott, 2011) or because they began with lower expectations of condom use (see also Broaddus et al., 2010). By selecting strongly assertive strategies (like Withholding Sex), WSM may feel better able to defend against coercive strategies to have condomless sex.

This is the first time, to our knowledge, that condom insistence strategies have been examined in an MSM sample, and it is noteworthy that MSM displayed a different response pattern than either MSW or WSM in regards to condom negotiation. MSM were more likely to select the strategies that used a Relationship Conceptualization or Deceptive strategy. This suggests that MSM have a different approach to condom negotiation. Their approach may reflect the more flexible, or complex, power dynamic during condom negotiation for MSM, where either partner could be the insertive partner (or “top”) and wear a condom. Because of this, among MSM neither partner is likely to have been primarily socialized in situations where condom use always depends on negotiating with a male partner for him to use a condom (something that WSM experience; Noar et al., 2002) or in situations where they are the only partner who could/should wear a condom (as MSW experience). This may contribute to the difference in condom insistence strategies selected, compared with WSM and MSW.

In the context of the MSM sample, the Deception strategy may not necessarily be used in order to “fool” a partner into using a condom, but instead may be invoked as a means to remind partners that condoms should be used and to establish themselves as a condom user. MSM are a known and relatively self-aware priority population for HIV intervention, and this cultural climate may influence their preferences for condom insistence strategies. Thus, MSM who are more risk averse may still wish to insist on condom use, even with a partner who claims to be HIV negative or claims to have an undetectable viral load (LeBlanc et al., 2014). The use of strategies like Relationship Conceptualizing or Deception may be preferred by MSM because these strategies are more normalizing and less confrontational than Withholding Sex or reminding a partner about the risks of HIV/STI transmission.

Adam (2006) pointed out that people who are at higher risk for HIV (like MSM) must navigate a more complex reality, due to the competing ideas around safer sex practices, sexual intimacy, and pleasure, and both personal and community responsibility to prevent HIV transmission. Many MSM experience a sense of altruism and community, and apply this to their moral reasoning about sexual safety (Cristian Rangel & Adam, 2014; Davis, 2008). This sentiment may also underlie the stronger correlation between MSM’s own interest in condom use and their perception of the hypothetical partner’s interest. The MSM in the current study may have a higher expectation that their partners will be interested in condom use because they see themselves and their partners as members of the same community, with shared goals.

Relationship Motivation

The associations between relationship motivation and sexual risk-taking intentions were at times in the directions predicted, and at times contradictory. For instance, HRM was associated with increased stated interest in condom use. This finding seems in conflict with the idea that HRM individuals would be willing to forgo condom use to facilitate intimacy. However, individuals who are strongly motivated by relationship goals are eager to present themselves as attractive partners (Hammer et al., 1996). In the current context, this could involve portraying themselves as individuals who care about their partners’ sexual safety and as being low risk for STI transmission.

Contrary to Hypothesis 5, both groups of LRM men reported a significantly greater willingness to engage in condomless sex than HRM men. In contrast, in agreement with our predictions, in WSM it was HRM WSM who showed greater intentions for condomless sex, when they also judged the hypothetical partner as more familiar. This suggests that relationship motivation may function differently in men than in women, and that the mechanisms may be more complex than anticipated. It may be that the LRM men showed an increased willingness because they perceived themselves and the hypothetical partner as low risk for negative outcomes (Comer & Nemeroff, 2000; Downing-Matibag & Geisinger, 2009). LRM (and especially male) participants who view their risk of STI transmission as low may also focus more on obtaining sexual pleasure than protecting themselves (or their partner), given the transitory nature that they may ascribe to such encounters.

It is also possible that, since male participants had higher expectations for partner interest in condom use, when those with HRM were presented with a partner who was resistant to condom use, this was more alarming. Individuals who are more strongly motivated by relationship goals are more sensitive to risk cues (Zawacki et al., 2009). Thus, the hypothetical partner’s unexpected resistance to condom use was likely perceived as an additional risk cue for men. In contrast WSM, who tended to begin with a lower expectation of partner interest in condom use, may not have viewed a resistant reaction as an additional risk cue.

Further research is needed to more deeply explore the observed gender differences in the association between relationship motivation and sexual risk-taking behavior, as well as the impact of partner familiarity. An examination of the responses of the subsample of participants who responded to the familiarity check item indicates that partner familiarity may have some impact on an individual’s perceptions and behavioral intentions during a sexual encounter. For instance, participants who rated the hypothetical partner as less familiar also had higher stated interest in condoms. Thus, a greater sense of familiarity may influence how we perceive sexual risk taking: new partners who feel more familiar may be seen as a lower risk for STI transmission (see also Sparling & Cramer, 2015). It may also be that those who are more cautious and avoid sexual risk taking are also slower to warm to a new potential partner and have stricter criteria for developing a sense of familiarity. Though these results should be interpreted with caution, since only a subsample of the participants received the familiarity item, the current findings do suggest fruitful future avenues of investigation to replicate these findings and further explore the potential role of partner familiarity in sexual risk-taking behavior.

Limitations of the Study

There are limitations to be considered. We did not include samples of women who have sex with women, or any other gender/sexuality minority groups (besides MSM). This was the result of purposeful methodological exclusion criteria in order to focus on sexual decision making and condom negotiation under conditions where at least one partner could use a condom on a flesh penis. However, this limits the generalizability of our findings, as does the fact that our sample was primarily White. Additionally, the subsample of respondents who responded to the familiarity item did not contain a sufficient number of MSM to make any inferences about how the familiarity of new sex partners may influence condom use in this population. Future work should explore how this factor may impact sexual health decision making in MSM and should explore these questions in more diverse samples.

Additionally, due to the limitations of the study methods, we were not able to determine how many participants selected a particular response option from the condom insistence strategy array simply because it was the closest approximation of their ideal choice. For example, MSM may have preferred the “deceptive” response (“I would tell Chris that I always have sex with condoms [even though sometimes I don’t]”) because it was the closest approximation to their ideal choice, which could have been non-deceptive. The constraints of our study methodology also did not allow us to successfully pursue alternatives to dichotomizing the RMS variable for some analyses. Splitting the data in this way resulted in some analytical limitations, since this approach excluded participants at the midpoint (n = 13) and created an artificial separation between participants who scored close to the midpoint. The method of measuring relationship motivation itself also presents a limitation in the current study. This scale asks participants to respond regarding their typical motivations to pursue or maintain a romantic relationship, but it is unknown how well this may have mapped onto their actual intentions during the vignette. More work is needed to untangle how individuals’ relationship motivation might be expressed differently during encounters with different partners. Regardless, the current findings represent foundational work on the concept of relationship motivation and demonstrate that this construct is worthy of further investigation in the context of sexual health decision making.

Finally, when using a hypothetical scenario it is difficult to fully capture all of the elements present in reality (Hughes & Huby, 2004); thus, a vignette may not invoke the same degree of emotional response (Collett & Childs, 2011). However, the current study does demonstrate participants’ sexual risk-taking intentions, which has been shown to be predictive of actual behavior (e.g., Sheeran, Abraham, & Orbell, 1999; Sheeran & Orbell, 1998; Turchik & Gidycz, 2012).

Conclusion

The results of this study demonstrate that many factors can influence an individual’s sexual risk-taking intentions during condom negotiation with a new sexual partner. MSW showed significantly greater sexual risk-taking intentions, particularly if they had lower relationship motivation. Conversely, WSM showed greater sexual risk-taking intentions if they had higher relationship motivation and also viewed their partner as more familiar. MSM took a different approach to condom negotiation than either the MSW or WSM; they selected strategies that were more in line with maintaining the mood of the sexual encounter, while still asserting a need for condom use. These results indicate that there are a number of conditions that may make it more difficult to recognize risk during a sexual encounter, including relationship motivation and partner familiarity. It is particularly striking that WSM had lower expectations that their partner would be interested in condom use and thus may have failed to view the partner’s resistance to condom use as an additional risk cue (unlike both male samples). Future work should explore how shifting social expectations for safer sex and changing attitudes toward safer sex practices, like condom use, can foster a view that one need not sacrifice enjoyment for safety and to combat the incorrect heuristic that condoms always reduce pleasure (see Wood et al., 2018) or interfere with the establishment of intimate relationships.

Disclosure statement

The authors declare no potential conflict of interest.

Notes

¹ Note that sexual preference was used in the current study as a proxy for sexual orientation, as this has been identified as the most critical aspect of orientation identity (see Sell, 2007). Note also that the terms “MSM,” “WSM,” and “MSW” are used in this study because the focus of the methodology was on the sexual behavior of participants. See Young and Meyer (2005) for a discussion on the limitations of using such terms to identify sexual minority groups.

² At the request of our reviewers we re-ran these analyses, testing for an interaction between gender/sexuality group and relationship motivation. The interaction term was not significant.

³ At the request of our reviewers, we conducted a sensitivity analysis and determined that our results held after excluding those in monogamous sexual relationships; thus, these participants were retained for all analyses.