JOURNAL OF SEX RESEARCH.50(1).37-47.2013 Copyright The Society for the Scientific Study of Sexuality Routledge ISSN:0022-4499 print/1559-8519 online Taylor Francis Group D0L:10.1080/00224499.2011.623797 Identifying and Explicating Variation among Friends with Benefits Relationships Paul A.Mongeau,Kendra Knight,and Jade Williams Hugh Downs School of Human Communication,Arizona State University Jennifer Eden Department of Communication,Northern Illinois University Christina Shaw Hugh Downs School of Human Communication,Arizona State University This two-study report identifies and validates a typology containing seven types of"friends with benefits relationships"(FWBRs).Study I asked heterosexual students to define the term FWBR and to describe their experience with the relationship type.Qualitative analysis of these data identified seven types of FWBRs (true friends,network opportunism,just sex, three types of transition in [successful,failed,and unintentionalJ,and transition out).Study 2 quantitatively differentiates these relationship types in the amount of nonsexual interaction, strength of the friendship at the first sexual interaction,and history of romantic relationships with the FWBR partner (before the FWBR,after it,or both).Results from both studies clearly suggest that FWBRs represent a diverse set of relationship formulations where both the benefits (ie.,repeated sexual contact)and the friends (i.e.,relationship between part- ners)vary widely.In many cases,FWBRs represent a desire for,or an attempt at,shifting the relationship from friends to a romantic partnership.Other implications are discussed,as are limitations and directions for future research.The diverse nature of FWBRs provides challenges for researchers that likely require multiple methods and theoretical frames. One of the few constants surrounding heterosexual the predominant sexual standard was "permissiveness courtship in the United States is change,as each gener- with affection."where sexual interaction was acceptable ation alters premarital romantic and sexual norms and if and only when partners were firmly committed to one practices(Bailey,1988;Wells Twenge,2005).Whether another(Perlman Sprecher,in press;Sprecher,1989). the result of a sexual revolution or a series of more Campus sexual standards in the 21st century's first gradual evolutions(Bailey,1999),the past half-century decade are quite permissive (Bogle,2008;Perlman witnessed drastic shifts in premarital sexual attitudes Sprecher,in press)and center on "hookups,"which and behaviors (Wells Twenge,2005).For example, are typically defined as strangers or acquaintances who in the 1950s and early 1960s,the predominant sexual engage in sexual interaction without expecting future standard was abstinence.where intercourse was reserved interaction (e.g.,Bogle,2008;Paul Hayes,2002).This for marriage(Perlman Sprecher,in press).A sexually investigation focuses on identifying and explicating charged campus tradition of the day was the "panty the nature of "friends with benefits relationships" raid,"where men would storm female dormitories,rifle (FWBRs),a permissive sexual practice closely related through dresser drawers,steal coeds'lingerie,and to hookups.In FWBRs,friends,who are not in a proudly display the loot (Bailey,1999).In the 1970s, romantic relationship,engage in multiple sexual interac- tions without the expectation that those interactions Study I data were presented to the National Communication reflect romantic intents or motivations (Epstein,Calzo Association meeting in Chicago,IL (November 2007),and Study 2 Smiler,Ward,2009).Given our definition,FWBRs data were presented at the National Communication Association meet- ing in San Diego,CA (November 2008).We thank Artemio Ramirez differ from hookups in two ways:First,FWBRs are for his help in Study I data collection,as well as Jason Peterson and more likely than hookups to occur between friends Merideth Bruck for their help in the Study I data analysis. Thus,FWBRs likely create expectations of more non- Correspondence should be addressed to Paul A.Mongeau,Hugh sexual interaction than do hookups.Second,sexual Downs School of Human Communication,Arizona State University interaction in FWBRs,more than in hookups,is likely P.O.Box 871205,Tempe,AZ 85287-1205.E-mail:paul.mongeau@ to be repeated. asu.edu
Identifying and Explicating Variation among Friends with Benefits Relationships Paul A. Mongeau, Kendra Knight, and Jade Williams Hugh Downs School of Human Communication, Arizona State University Jennifer Eden Department of Communication, Northern Illinois University Christina Shaw Hugh Downs School of Human Communication, Arizona State University This two-study report identifies and validates a typology containing seven types of ‘‘friends with benefits relationships’’ (FWBRs). Study 1 asked heterosexual students to define the term FWBR and to describe their experience with the relationship type. Qualitative analysis of these data identified seven types of FWBRs (true friends, network opportunism, just sex, three types of transition in [successful, failed, and unintentional], and transition out). Study 2 quantitatively differentiates these relationship types in the amount of nonsexual interaction, strength of the friendship at the first sexual interaction, and history of romantic relationships with the FWBR partner (before the FWBR, after it, or both). Results from both studies clearly suggest that FWBRs represent a diverse set of relationship formulations where both the benefits (i.e., repeated sexual contact) and the friends (i.e., relationship between partners) vary widely. In many cases, FWBRs represent a desire for, or an attempt at, shifting the relationship from friends to a romantic partnership. Other implications are discussed, as are limitations and directions for future research. The diverse nature of FWBRs provides challenges for researchers that likely require multiple methods and theoretical frames. One of the few constants surrounding heterosexual courtship in the United States is change, as each generation alters premarital romantic and sexual norms and practices (Bailey, 1988; Wells & Twenge, 2005). Whether the result of a sexual revolution or a series of more gradual evolutions (Bailey, 1999), the past half-century witnessed drastic shifts in premarital sexual attitudes and behaviors (Wells & Twenge, 2005). For example, in the 1950s and early 1960s, the predominant sexual standard was abstinence, where intercourse was reserved for marriage (Perlman & Sprecher, in press). A sexually charged campus tradition of the day was the ‘‘panty raid,’’ where men would storm female dormitories, rifle through dresser drawers, steal coeds’ lingerie, and proudly display the loot (Bailey, 1999). In the 1970s, the predominant sexual standard was ‘‘permissiveness with affection,’’ where sexual interaction was acceptable if and only when partners were firmly committed to one another (Perlman & Sprecher, in press; Sprecher, 1989). Campus sexual standards in the 21st century’s first decade are quite permissive (Bogle, 2008; Perlman & Sprecher, in press) and center on ‘‘hookups,’’ which are typically defined as strangers or acquaintances who engage in sexual interaction without expecting future interaction (e.g., Bogle, 2008; Paul & Hayes, 2002). This investigation focuses on identifying and explicating the nature of ‘‘friends with benefits relationships’’ (FWBRs), a permissive sexual practice closely related to hookups. In FWBRs, friends, who are not in a romantic relationship, engage in multiple sexual interactions without the expectation that those interactions reflect romantic intents or motivations (Epstein, Calzo, Smiler, & Ward, 2009). Given our definition, FWBRs differ from hookups in two ways: First, FWBRs are more likely than hookups to occur between friends. Thus, FWBRs likely create expectations of more nonsexual interaction than do hookups. Second, sexual interaction in FWBRs, more than in hookups, is likely to be repeated. Study 1 data were presented to the National Communication Association meeting in Chicago, IL (November 2007), and Study 2 data were presented at the National Communication Association meeting in San Diego, CA (November 2008). We thank Artemio Ramirez for his help in Study 1 data collection, as well as Jason Peterson and Merideth Bruck for their help in the Study 1 data analysis. Correspondence should be addressed to Paul A. Mongeau, Hugh Downs School of Human Communication, Arizona State University, P.O. Box 871205, Tempe, AZ 85287-1205. E-mail: paul.mongeau@ asu.edu JOURNAL OF SEX RESEARCH, 50(1), 37–47, 2013 Copyright # The Society for the Scientific Study of Sexuality ISSN: 0022-4499 print=1559-8519 online DOI: 10.1080/00224499.2011.623797
MONGEAU,KNIGHT.WILLIAMS.EDEN.AND SHAW College students understand the sexual scripts sur- 2009;Reeder,2000).Thus,strings cannot differentiate rounding both hookups and FWBRs (Epstein et al., FWBRs from romantic relationships because many 2009).According to script theory,"sexuality is learned cases,as we attempt to demonstrate,lie between from culturally available'sexual scripts'that define what emotion-free FWBRs and emotion-laden romantic counts as sex,how to recognize sexual situations,and relationships what to do in relational and sexual encounters"(Kim et al.,2007,p.146;see also Gagnon Simon,1973). Friends with Benefits Epstein et al.demonstrated substantial variation in how hookups occur,but provided no data on FWBRs. A second reason why FWBRs vary is the nature of Our reading of the literature suggests similar variation partners'friendships.The FWBR literature describes among FWBRs (e.g.,Bisson Levine,2009;Epstein variation in benefits (from only kissing to oral/vaginal et al.,2009;Hughes,Morrison,Asada,2005; intercourse:Bisson Levine,2009):however,the nature Lehmiller,VanderDrift,Kelly,2010;Mongeau, of the friends has gone unquestioned.Mongeau et al. Ramirez,Vorell,2003).As the FWBR label likely cov- (2003),however,indicated that FWBR partners differ ers(and obscures)a variety of relational types,the extant in how well they know each other.Consistent with the FWBR literature lacks depth.Therefore,the primary common definition,many FWBR partners know each goals of this investigation were to review the literature other well and care for each other before initiating with a specific focus on how FWBRs vary,to identify sexual contact(Reeder,2000),perhaps allowing them types of FWBRs in students'descriptions and definitions to investigate romantic potential (Bleske Buss, (Study 1),and to validate those types by demonstrating 2000).On the other hand,some FWBR partners initiate that they differ systematically(Study 2). sexual interaction soon after initially meeting(Knight, 2008;Mongeau et al.,2003).In addition,some FWBRs include romantic history.Some FWBRs morph into Variety among Friends with Benefits Relationships romantic relationships(Bisson Levine,2009),whereas (FWBRs) others represent the "smoldering embers"of a past romantic relationship (e.g.,Mongeau et al.,2003,p.19). The extant literature assumes that FWBRs are a In summary,given the variety we see in the FWBR singular relationship type.At the same time,however, literature,this two-study report attempted to identify it provides evidence of variation in both the presence and explicate variation among FWBRs.Specifically, of romantic motivations and nature of the friendship Study 1 attempted to identify different types of FWBRs. (e.g.,Bisson Levine,2009;Hughes et al.,2005;Leh- Study 2,in turn,attempted to validate these types by miller et al.,2010;Mongeau et al.,2003).We consider investigating differences in pre-sexual interaction, each characteristic in turn. friendship strength,and romantic history. The Nature of "Strings" Study 1 Ideally,FWBRs are simple:Friends have sex repeat- Method edly with "no strings attached"(e.g.,Bisson Levine, 2009;Epstein et al.,2009;Hughes et al.,2005;Levine Participants and procedures. Participants included Mongeau,2010).The absence of strings suggests a undergraduate students in communication classes at lack of romantic ties,motivations,or expectations that two very large public U.S.universities (one Southwest- restrict extra-dyadic behavior (Hughes et al.,2005). ern,one Midwestern),who received extra credit for their Friends add sex to an existing friendship (Knight's, participation.The Southwestern sample included 177 2008,"add sex and stir"approach)to avoid drama participants,predominately women (n=111;62.7%). inherent in romantic relationships. Although ethnicity and age were not measured,past sam- In many cases,the reality of FWBRs is actually ples from this pool were predominately Whites of typical quite complex.Some FWBR partners do have romantic college age.The Midwestern sample included 102 parti- feelings.When romantic interest is mutual,a FWBR can cipants (39 males,61 females,and two with no report; become a romantic relationship (Afifi Faulkner,2000; Mage=20.47,SD=3.52).Participants self-identified as Bisson Levine,2009).Conversely,unreciprocated Caucasian (81.4%),African American (9.8%),Asian romantic feelings are common in FWBRs and generate American (3.9%),Hispanic American(2%),and "other" discomfort (Bisson Levine.2009:Hughes et al..2005: (1%).Over one-half of the participants(51.4%)reported Mongeau et al.,2003),perhaps leading to dissolution personal experiences with FWBRs.Approximately one because one partner wanted something more(e.g.,a real in eight participants(11.9%)indicated being in a FWBR relationship;Knight,2008).In other cases,FWBRs rep- at the time of the data collection. resent a compromise,with one partner waiting for the Methods were approved by both campuses' other to develop romantic feelings (Epstein et al., institutional review boards.Students in upper-and 38
College students understand the sexual scripts surrounding both hookups and FWBRs (Epstein et al., 2009). According to script theory, ‘‘sexuality is learned from culturally available ‘sexual scripts’ that define what counts as sex, how to recognize sexual situations, and what to do in relational and sexual encounters’’ (Kim et al., 2007, p. 146; see also Gagnon & Simon, 1973). Epstein et al. demonstrated substantial variation in how hookups occur, but provided no data on FWBRs. Our reading of the literature suggests similar variation among FWBRs (e.g., Bisson & Levine, 2009; Epstein et al., 2009; Hughes, Morrison, & Asada, 2005; Lehmiller, VanderDrift, & Kelly, 2010; Mongeau, Ramirez, & Vorell, 2003). As the FWBR label likely covers (and obscures) a variety of relational types, the extant FWBR literature lacks depth. Therefore, the primary goals of this investigation were to review the literature with a specific focus on how FWBRs vary, to identify types of FWBRs in students’ descriptions and definitions (Study 1), and to validate those types by demonstrating that they differ systematically (Study 2). Variety among Friends with Benefits Relationships (FWBRs) The extant literature assumes that FWBRs are a singular relationship type. At the same time, however, it provides evidence of variation in both the presence of romantic motivations and nature of the friendship (e.g., Bisson & Levine, 2009; Hughes et al., 2005; Lehmiller et al., 2010; Mongeau et al., 2003). We consider each characteristic in turn. The Nature of ‘‘Strings’’ Ideally, FWBRs are simple: Friends have sex repeatedly with ‘‘no strings attached’’ (e.g., Bisson & Levine, 2009; Epstein et al., 2009; Hughes et al., 2005; Levine & Mongeau, 2010). The absence of strings suggests a lack of romantic ties, motivations, or expectations that restrict extra-dyadic behavior (Hughes et al., 2005). Friends add sex to an existing friendship (Knight’s, 2008, ‘‘add sex and stir’’ approach) to avoid drama inherent in romantic relationships. In many cases, the reality of FWBRs is actually quite complex. Some FWBR partners do have romantic feelings. When romantic interest is mutual, a FWBR can become a romantic relationship (Afifi & Faulkner, 2000; Bisson & Levine, 2009). Conversely, unreciprocated romantic feelings are common in FWBRs and generate discomfort (Bisson & Levine, 2009; Hughes et al., 2005; Mongeau et al., 2003), perhaps leading to dissolution because one partner wanted something more (e.g., a real relationship; Knight, 2008). In other cases, FWBRs represent a compromise, with one partner waiting for the other to develop romantic feelings (Epstein et al., 2009; Reeder, 2000). Thus, strings cannot differentiate FWBRs from romantic relationships because many cases, as we attempt to demonstrate, lie between emotion-free FWBRs and emotion-laden romantic relationships. Friends with Benefits A second reason why FWBRs vary is the nature of partners’ friendships. The FWBR literature describes variation in benefits (from only kissing to oral=vaginal intercourse; Bisson & Levine, 2009); however, the nature of the friends has gone unquestioned. Mongeau et al. (2003), however, indicated that FWBR partners differ in how well they know each other. Consistent with the common definition, many FWBR partners know each other well and care for each other before initiating sexual contact (Reeder, 2000), perhaps allowing them to investigate romantic potential (Bleske & Buss, 2000). On the other hand, some FWBR partners initiate sexual interaction soon after initially meeting (Knight, 2008; Mongeau et al., 2003). In addition, some FWBRs include romantic history. Some FWBRs morph into romantic relationships (Bisson & Levine, 2009), whereas others represent the ‘‘smoldering embers’’ of a past romantic relationship (e.g., Mongeau et al., 2003, p. 19). In summary, given the variety we see in the FWBR literature, this two-study report attempted to identify and explicate variation among FWBRs. Specifically, Study 1 attempted to identify different types of FWBRs. Study 2, in turn, attempted to validate these types by investigating differences in pre-sexual interaction, friendship strength, and romantic history. Study 1 Method Participants and procedures. Participants included undergraduate students in communication classes at two very large public U.S. universities (one Southwestern, one Midwestern), who received extra credit for their participation. The Southwestern sample included 177 participants, predominately women (n ¼ 111; 62.7%). Although ethnicity and age were not measured, past samples from this pool were predominately Whites of typical college age. The Midwestern sample included 102 participants (39 males, 61 females, and two with no report; Mage ¼ 20.47, SD ¼ 3.52). Participants self-identified as Caucasian (81.4%), African American (9.8%), Asian American (3.9%), Hispanic American (2%), and ‘‘other’’ (1%). Over one-half of the participants (51.4%) reported personal experiences with FWBRs. Approximately one in eight participants (11.9%) indicated being in a FWBR at the time of the data collection. Methods were approved by both campuses’ institutional review boards. Students in upper- and MONGEAU, KNIGHT, WILLIAMS, EDEN, AND SHAW 38
VARIATION AMONG FRIENDS WITH BENEFITS lower-division communication classes were invited to expanded.Discussion resulted in identification,labeling, participate in a study on current campus dating norms. and definition of FWBR types.Seven relational types The FWBR term was not used in the study invitation emerged:true friends,just sex,network opportunism, materials.Students were given a copy of the survey, transition in (including successful,failed,and uninten- instructed to complete it outside of class,and asked to tional),and transition out. return it to their instructor. The FWBR section of the survey first asked parti- Results cipants to define FWBRs in their own words.On a sub- sequent page,FWBRs were defined as follows: Although participants described their FWBRs in myriad ways,the one (and only)point of agreement involveing]platonic friends (i.e.,those not involved in a was sexual activity.That being said,however,the nature romantic relationship)who engage in some degree of of the activity varied across responses.Some responses sexual intimacy on multiple occasions.This sexual described intercourse (oral or vaginal).whereas others activity could range from kissing to sexual intercourse described less intimate sexual activity (e.g.,kissing)or and is a repeated part of your friendship such that it is used ambiguous terms (e.g.,making out).Beyond sex, not just a one-night stand. responses markedly diverged on emotional investment, communication (intra-and interdyadic),secrecy,exclu- Following the definition,participants reported their sivity,obligations,investment,and dating. experience with FWBRs and,if they had any,described Seven types of FWBRs.Analyses of participants' how sexual involvement began,how the FWBR differed responses identified seven distinguishable FWBR types from a "typical"romantic relationship,and (if it had that differed in social,interactive,and relational charac- ended)why it ended. teristics.Specifically,these types differed in the nature of the relationship and interactions between partners. Data Analysis including history of,or desire for,romantic relation- ships.The types include true FWBRs,just sex,network The analytic plan for Study 1 was emergent.Follow- opportunism,three types of transition in (successful, ing Mongeau,Jacobsen,and Donnerstein's(2007)work failed,and unintentional),and transition out.Each of on dates,we originally attempted a content analysis these types is described,and an example provided,in to identify FWBRs'essential characteristics.Analysis turn. began in an emic fashion (i.e.,categories emerged from The true friends type reflects the traditional FWBR the data,as no previous typology existed;Lindlof definition (i.e.,close friends who have sex on multiple Taylor,2002).Initially,two authors and a research occasions;e.g.,Bisson Levine,2009;Epstein et al., assistant jointly coded several pages of data to establish 2009).Participants express love,trust,and respect for consistency using thought-turns as the unit of analysis. an important friend who is considered a safe sexual or data "chunks"(Lindlof Taylor,2002,p.219). partner (Levine Mongeau,2010).Partners frequently Through constant comparison,each chunk was either interact in varied contexts.True friends appear similar given an existing code or a new code was developed. to,but are not labeled as,romantic relationships.For We then met weekly to discuss subsequent coding. example,one participant defined FWBRs as follows: Despite many weeks of refining and redefining catego- "It means someone who you know and care about as ries,we could not reach acceptable intercoder reliability. a friend/person who you also happen to have a sexual In considering our inability to achieve reliability,we relationship with"(Participant 039B) uncovered a flaw in a guiding assumption.Following Conversely,just sex partners interact almost exclus- the literature of the day,we had assumed that FWBRs ively to arrange and carry out sexual interaction.Other represented a singular relationship type such that open than sexual encounters.little interaction occurs between and axial coding should identify essential characteristics. partners.In these cases,the "friend"in FWBR is a mis- In (re)reading the data,variation across responses nomer as partners engage,essentially,in serial hookups became clear that violated the singularity assumption. (Paul Hayes,2002).One participant defined FWBRs For example,some participants described close,intimate as follows:"You don't really care about the person in friendships,whereas others referred to FWBRs as inter- a special way,but s/he is just there when you are feeling actions between strangers.Rejecting the singularity sexual"(Participant 039A). assumption led to our second analytic phase-the Third,network opportunism involves sexual interac- identification of FWBR types. tion between friends (although not particularly close The first three authors returned to the data as a ones)who share network links.Those shared links allow whole,and independently developed typologies contain- partners to interact,typically while consuming alcohol. ing four to six FWBR types.Discussion revealed con- These partners engage in sexual activity if neither has siderable overlap among typologies,and disagreements found a different sexual partner for the night.In short, were resolved such that categories collapsed and these partners act as a sexual fail-safe.Given the 39
lower-division communication classes were invited to participate in a study on current campus dating norms. The FWBR term was not used in the study invitation materials. Students were given a copy of the survey, instructed to complete it outside of class, and asked to return it to their instructor. The FWBR section of the survey first asked participants to define FWBRs in their own words. On a subsequent page, FWBRs were defined as follows: involve[ing] platonic friends (i.e., those not involved in a romantic relationship) who engage in some degree of sexual intimacy on multiple occasions. This sexual activity could range from kissing to sexual intercourse and is a repeated part of your friendship such that it is not just a one-night stand. Following the definition, participants reported their experience with FWBRs and, if they had any, described how sexual involvement began, how the FWBR differed from a ‘‘typical’’ romantic relationship, and (if it had ended) why it ended. Data Analysis The analytic plan for Study 1 was emergent. Following Mongeau, Jacobsen, and Donnerstein’s (2007) work on dates, we originally attempted a content analysis to identify FWBRs’ essential characteristics. Analysis began in an emic fashion (i.e., categories emerged from the data, as no previous typology existed; Lindlof & Taylor, 2002). Initially, two authors and a research assistant jointly coded several pages of data to establish consistency using thought-turns as the unit of analysis, or data ‘‘chunks’’ (Lindlof & Taylor, 2002, p. 219). Through constant comparison, each chunk was either given an existing code or a new code was developed. We then met weekly to discuss subsequent coding. Despite many weeks of refining and redefining categories, we could not reach acceptable intercoder reliability. In considering our inability to achieve reliability, we uncovered a flaw in a guiding assumption. Following the literature of the day, we had assumed that FWBRs represented a singular relationship type such that open and axial coding should identify essential characteristics. In (re)reading the data, variation across responses became clear that violated the singularity assumption. For example, some participants described close, intimate friendships, whereas others referred to FWBRs as interactions between strangers. Rejecting the singularity assumption led to our second analytic phase—the identification of FWBR types. The first three authors returned to the data as a whole, and independently developed typologies containing four to six FWBR types. Discussion revealed considerable overlap among typologies, and disagreements were resolved such that categories collapsed and expanded. Discussion resulted in identification, labeling, and definition of FWBR types. Seven relational types emerged: true friends, just sex, network opportunism, transition in (including successful, failed, and unintentional), and transition out. Results Although participants described their FWBRs in myriad ways, the one (and only) point of agreement was sexual activity. That being said, however, the nature of the activity varied across responses. Some responses described intercourse (oral or vaginal), whereas others described less intimate sexual activity (e.g., kissing) or used ambiguous terms (e.g., making out). Beyond sex, responses markedly diverged on emotional investment, communication (intra- and interdyadic), secrecy, exclusivity, obligations, investment, and dating. Seven types of FWBRs. Analyses of participants’ responses identified seven distinguishable FWBR types that differed in social, interactive, and relational characteristics. Specifically, these types differed in the nature of the relationship and interactions between partners, including history of, or desire for, romantic relationships. The types include true FWBRs, just sex, network opportunism, three types of transition in (successful, failed, and unintentional), and transition out. Each of these types is described, and an example provided, in turn. The true friends type reflects the traditional FWBR definition (i.e., close friends who have sex on multiple occasions; e.g., Bisson & Levine, 2009; Epstein et al., 2009). Participants express love, trust, and respect for an important friend who is considered a safe sexual partner (Levine & Mongeau, 2010). Partners frequently interact in varied contexts. True friends appear similar to, but are not labeled as, romantic relationships. For example, one participant defined FWBRs as follows: ‘‘It means someone who you know and care about as a friend=person who you also happen to have a sexual relationship with’’ (Participant 039B). Conversely, just sex partners interact almost exclusively to arrange and carry out sexual interaction. Other than sexual encounters, little interaction occurs between partners. In these cases, the ‘‘friend’’ in FWBR is a misnomer as partners engage, essentially, in serial hookups (Paul & Hayes, 2002). One participant defined FWBRs as follows: ‘‘You don’t really care about the person in a special way, but s=he is just there when you are feeling sexual’’ (Participant 039A). Third, network opportunism involves sexual interaction between friends (although not particularly close ones) who share network links. Those shared links allow partners to interact, typically while consuming alcohol. These partners engage in sexual activity if neither has found a different sexual partner for the night. In short, these partners act as a sexual fail-safe. Given the VARIATION AMONG FRIENDS WITH BENEFITS 39
MONGEAU,KNIGHT.WILLIAMS.EDEN.AND SHAW common social network,network opportunism likely discussion of hookups).Thus,the primary goal of Study involves more interaction than just sex.but with less I became to unpack the variety underlying students' breadth,depth,and frequency than true friends.One descriptions of FWBRs.Analyses suggested seven types participant described his or her FWBRs this way:"We of FWBRs(true friends,just sex,network opportunism, hung out and talked like normal friends but at the end transition in [successful,unintentional,and failed],and of the night (or party),we ended up in bed together transition out)that differ in friendship strength and instead of leaving each other"(Participant 066B). romantic history.The nature of romantic "strings"has Transition in FWBRs precede romantic relationship the most important implications for sexual and non- development (e.g.,Afifi Faulkner,2000;Bisson sexual interaction,so we discuss it briefly here. Levine,2009).Interactions in these FWBRs lead to,or reflect,romantic feelings.There is variation,however, FWBRs and romantic strings.In these data, among transition in cases.For example,some transition in FWBRs represent intentional attempts at romantic FWBRs are frequently juxtaposed with romantic rela- tionships.Given the normative nature of sexual interac- relationship initiation (some successful,others not).In other cases,the romantic transition appears to be an tion in very early relationship stages (e.g.,Bogle,2008; unintended byproduct of the sexual interaction. Wells Twenge,2005),some transition in FWBRs likely Given this variation,we divided the transition in type act as a bridge between platonic and romantic entangle- ments.Given that romantic relationship transitions into three parts:successful,unintentional,and failed. engender relational uncertainty (Mongeau,Serewicz, Successful transition in represents intentionally and effectively using a FWBR to initiate a romantic relation- Henningsen,Davis,2006;Solomon Knobloch, ship.For example,"I knew he was afraid of ruining the 2004),some FWBRs may facilitate uncertainty red- friendship,but I wanted more,and it worked.We are uction about the partner and his or her romantic poten- and have been a couple"(Participant 065A).Second, tial and interest.In this sense,FWBRs may serve an in unintentional transition in,a FWBR leads to a roman- investigative function previously fulfilled by first dates tic relationship,although it was not the respondent's (Mongeau,Serewicz,Therrien,2004). In addition,transition out FWBRs involve sexual original intent.For example.one participant described their FWBR as follows:"We didn't call ourself a couple interaction with a past romantic partner (i.e.,ex-sex). There are several potential advantages to ex-sex.First, but we were having sex just about everyday [sic ..Eventually it turned into us dating and actually participants considered their FWBR partner as a"safe" showing emotions for eachother [sic]instead of just hav- sexual partner (likely both in terms of safe-sex practices ing sex with nothing"(Participant 067C).Finally,in and not intentionally inflicting emotional or physical failed transition in,one or both partners attempted, pain).Second,ex-sex is likely familiar,both in terms unsuccessfully,to generate a romantic transition,but of the partner and his or her sexual(dis)likes that make interactions more predictable and,perhaps,enjoyable. continued sexual interactions.For example,one partici- pant said of their FWBR,"I wanted to make [the Third,sleeping with a former partner may be seen as relationship]more serious,he wanted to be single and advantageous to those desiring sexual interaction with- out increasing the number of lifetime sexual partners. not tied down"(Participant 096B). Finally,transition out FWBRs reflect sexual interac- Finally,fanning sexual flames might facilitate rekindling tions between partners from a terminated romantic partners'emotional connections. relationship.Partners no longer label their relationship Ex-sex appears in several recent investigations (e.g., as romantic,but continue sexual interactions:"We were Afifi Faulkner,2000;Dailey,Rossetto,Pfiester, Surra.2009:Koenig-Kellas,Bean.Cunningham, a couple,then I broke up with her but we continued the FWB kind of relationship for about five more months" Cheng,2008;Smith Morrison,2010).Dailey et al. (Participant 019A). suggested that some dating relationships are intermittent (i.e.,on again-off again)such that FWBRs might represent an intermediate position between exclusively Study 1 Discussion dating and totally terminated.Again,this suggests that Consistent with the literature of the time(e.g.,Bisson some FWBRs represent an intermediate position between friendships and romantic interactions. Levine,2009;Hughes et al.,2005;Mongeau et al., 2003),we began this project assuming that FWBRs represented a single relationship type.Numerous read- Utility and implications of the FWBR label.Given ings of our data led us to reject the unitary assumption our results,the FWBR label covers (obscures)a variety and to an alternative reading of the literature.Thus, of relationship genres.The wide swath of relationships although college students have a common understand- called FWBRs parallels the strategic use of hookup to ing of FWBRs (e.g.,Bisson Levine,2009;Epstein describe nearly any sexual interaction in nearly any rela- et al.,2009;Hughes et al.,2005),actual practice reflects tional context(Epstein et al.,2009;Paul Hayes,2002). substantial variation (cf.the Epstein et al.,2009, The term hookup says a lot(i.e.,some sexual interaction 40
common social network, network opportunism likely involves more interaction than just sex, but with less breadth, depth, and frequency than true friends. One participant described his or her FWBRs this way: ‘‘We hung out and talked like normal friends but at the end of the night (or party), we ended up in bed together instead of leaving each other’’ (Participant 066B). Transition in FWBRs precede romantic relationship development (e.g., Afifi & Faulkner, 2000; Bisson & Levine, 2009). Interactions in these FWBRs lead to, or reflect, romantic feelings. There is variation, however, among transition in cases. For example, some transition in FWBRs represent intentional attempts at romantic relationship initiation (some successful, others not). In other cases, the romantic transition appears to be an unintended byproduct of the sexual interaction. Given this variation, we divided the transition in type into three parts: successful, unintentional, and failed. Successful transition in represents intentionally and effectively using a FWBR to initiate a romantic relationship. For example, ‘‘I knew he was afraid of ruining the friendship, but I wanted more, and it worked. We are and have been a couple’’ (Participant 065A). Second, in unintentional transition in, a FWBR leads to a romantic relationship, although it was not the respondent’s original intent. For example, one participant described their FWBR as follows: ‘‘We didn’t call ourself a couple but we were having sex just about everyday [sic]. ... Eventually it turned into us dating and actually showing emotions for eachother [sic] instead of just having sex with nothing’’ (Participant 067C). Finally, in failed transition in, one or both partners attempted, unsuccessfully, to generate a romantic transition, but continued sexual interactions. For example, one participant said of their FWBR, ‘‘I wanted to make [the relationship] more serious, he wanted to be single and not tied down’’ (Participant 096B). Finally, transition out FWBRs reflect sexual interactions between partners from a terminated romantic relationship. Partners no longer label their relationship as romantic, but continue sexual interactions: ‘‘We were a couple, then I broke up with her but we continued the FWB kind of relationship for about five more months’’ (Participant 019A). Study 1 Discussion Consistent with the literature of the time (e.g., Bisson & Levine, 2009; Hughes et al., 2005; Mongeau et al., 2003), we began this project assuming that FWBRs represented a single relationship type. Numerous readings of our data led us to reject the unitary assumption and to an alternative reading of the literature. Thus, although college students have a common understanding of FWBRs (e.g., Bisson & Levine, 2009; Epstein et al., 2009; Hughes et al., 2005), actual practice reflects substantial variation (cf. the Epstein et al., 2009, discussion of hookups). Thus, the primary goal of Study 1 became to unpack the variety underlying students’ descriptions of FWBRs. Analyses suggested seven types of FWBRs (true friends, just sex, network opportunism, transition in [successful, unintentional, and failed], and transition out) that differ in friendship strength and romantic history. The nature of romantic ‘‘strings’’ has the most important implications for sexual and nonsexual interaction, so we discuss it briefly here. FWBRs and romantic strings. In these data, FWBRs are frequently juxtaposed with romantic relationships. Given the normative nature of sexual interaction in very early relationship stages (e.g., Bogle, 2008; Wells & Twenge, 2005), some transition in FWBRs likely act as a bridge between platonic and romantic entanglements. Given that romantic relationship transitions engender relational uncertainty (Mongeau, Serewicz, Henningsen, & Davis, 2006; Solomon & Knobloch, 2004), some FWBRs may facilitate uncertainty reduction about the partner and his or her romantic potential and interest. In this sense, FWBRs may serve an investigative function previously fulfilled by first dates (Mongeau, Serewicz, & Therrien, 2004). In addition, transition out FWBRs involve sexual interaction with a past romantic partner (i.e., ex-sex). There are several potential advantages to ex-sex. First, participants considered their FWBR partner as a ‘‘safe’’ sexual partner (likely both in terms of safe-sex practices and not intentionally inflicting emotional or physical pain). Second, ex-sex is likely familiar, both in terms of the partner and his or her sexual (dis)likes that make interactions more predictable and, perhaps, enjoyable. Third, sleeping with a former partner may be seen as advantageous to those desiring sexual interaction without increasing the number of lifetime sexual partners. Finally, fanning sexual flames might facilitate rekindling partners’ emotional connections. Ex-sex appears in several recent investigations (e.g., Afifi & Faulkner, 2000; Dailey, Rossetto, Pfiester, & Surra, 2009; Koenig-Kellas, Bean, Cunningham, & Cheng, 2008; Smith & Morrison, 2010). Dailey et al. suggested that some dating relationships are intermittent (i.e., on again-off again) such that FWBRs might represent an intermediate position between exclusively dating and totally terminated. Again, this suggests that some FWBRs represent an intermediate position between friendships and romantic interactions. Utility and implications of the FWBR label. Given our results, the FWBR label covers (obscures) a variety of relationship genres. The wide swath of relationships called FWBRs parallels the strategic use of hookup to describe nearly any sexual interaction in nearly any relational context (Epstein et al., 2009; Paul & Hayes, 2002). The term hookup says a lot (i.e., some sexual interaction MONGEAU, KNIGHT, WILLIAMS, EDEN, AND SHAW 40
VARIATION AMONG FRIENDS WITH BENEFITS occurred)without saying what did or did not happen. In summary,Study I provides evidence of multiple Our suspicion is that the same is likely true of FWBR types suggested in the literature (e.g.,Bisson FWBR to the point where it appears difficult to deter- Levine,2009;Furman Hand,2006;Hughes et al., mine the difference between hookups and FWBRs. 2005;Mongeau et al.,2003).Not all participants,how- Hookups can involve relational partners (e.g.,friends ever,had experiences with FWBRs.Thus,the data and even romantic partners;Epstein et al.,2009), might contain both scripted(e.g.,definitions from part- and FWBRs can occur between relative strangers. icipants without FWBR experiences)and experiential Moreover,hookups (i.e.,sexual encounters)occur elements.Such a combination might make the data look within FWBRs.Thus,the use of both terms is likely more similar to the dominant script than if we had strategically ambiguous. included only experiential data.Moreover,given that The ambiguous use of FWBR may be useful in at we did not initially set out to identify FWBR types, least two contexts.First,"[U]sing a nonrelational label many questions remained.Therefore,Study 2 specifi- at first may be one way of dealing with the uncertainty cally investigated differences among,and frequency of, that comes in the first stages of dating"(Epstein et al., the seven FWBR types. 2009,p.421).Given the normative nature of non- relational sex on U.S.college campuses (Bogle,2008), relational implications of early sexual interactions are Study 2 likely unclear (e.g.,does it have romantic implications or was it a drunken hookup?).Using the FWBR label Two primary questions drove Study 2.The most might allow partners to avoid talking about their important goal was to validate the relational types.Spe- relationship and its trajectory,as the label does not need cifically,FWBR types should differ systematically in to be discussed or defined (Knight,2008).Second,the nonsexual interactions,romantic history,and friendship FWBR label might also be useful in on again-off again strength.Second,Study I methods did not allow us to relationships (Dailey et al.,2009),when it is unclear determine the frequency of the seven FWBR types. whether a couple is together or not.In both these cases, Therefore,we also considered how frequently each nonrelational labels such as FWBR may reflect(or gloss FWBR type occured (including gender differences in over)relational uncertainty or minimize perceptions of these reports). relational involvement to both the partner and the social network (Epstein et al.,2009). The FWBR label might also be useful in dealing Validating FWBR Types with the social network.For example,when a romantic Given our definitions,the seven FWBR types should transition was attempted but failed,the FWBR label differ in friendship strength,nonsexual interaction, may be superior to admitting that a couple tried dating, and romantic history.To validate our typology,Study 2 but it did not pan out.In such cases,the FWBR label focused on differences among FWBR types in these is used only retrospectively.The FWBR label can also variables.Small and/or nonsignificant differences across hide relational uncertainty from the social network just FWBR types would greatly hinder the validity and util- as it can with the partner.Calling a potentially budding ity of our typology. romantic relationship a FWBR might be superior to First,by definition,FWBRs types should differ in the admitting uncertainty about the relational definition, frequency of nonsexual interactions and friendship trajectory,and the partner's motivations.Finally, strength.Specifically,FWBR types characterized by the FWBR label might provide a socially appropriate closeness and trust (e.g.,true friends and successful tran- label for serial hookups with a stranger.Such com- sition in)should reflect more nonsexual interactions municative practices further blur the distinction and a stronger friendship at the initiation of sexual between relational and nonrelational sex (e.g.,Epstein activity,than marginally-related partners or those who etal.,2009). have contrasting relational goals(i.e.,just sex and failed Ambiguous use of the FWB label can be disadvan- transition in).Other FWB types(network opportunism tageous,as it can obscure differences in partners'desires and transition out)were expected to fall between these and expectations-that is,although two people extremes. explicitly agree to the FWBR label and appear to be Second,FWBR types should,by definition,differ in on the same page,they might actually think about the romantic experience.Participants in successful and relationship in fundamentally different ways.The com- unintentional transition in categories should report mon FWBR script(Epstein et al.,2009)could aggravate having a romantic relationship following their FWBR, such misunderstandings.For example,Partner A might whereas transition out partners should report a desire a "booty call,"whereas Partner B wants to move romantic relationship before the FWBR.Network toward a romantic relationship.In such cases,the per- opportunism,failed transition in,and(particularly)just son who wants "more"is probably at a distinct dis- sex FWBRs should be unlikely to include romantic advantage(Sprecher,Schmeeckle,Felmlee,2006). experiences 41
occurred) without saying what did or did not happen. Our suspicion is that the same is likely true of FWBR to the point where it appears difficult to determine the difference between hookups and FWBRs. Hookups can involve relational partners (e.g., friends and even romantic partners; Epstein et al., 2009), and FWBRs can occur between relative strangers. Moreover, hookups (i.e., sexual encounters) occur within FWBRs. Thus, the use of both terms is likely strategically ambiguous. The ambiguous use of FWBR may be useful in at least two contexts. First, ‘‘[U]sing a nonrelational label at first may be one way of dealing with the uncertainty that comes in the first stages of dating’’ (Epstein et al., 2009, p. 421). Given the normative nature of nonrelational sex on U.S. college campuses (Bogle, 2008), relational implications of early sexual interactions are likely unclear (e.g., does it have romantic implications or was it a drunken hookup?). Using the FWBR label might allow partners to avoid talking about their relationship and its trajectory, as the label does not need to be discussed or defined (Knight, 2008). Second, the FWBR label might also be useful in on again-off again relationships (Dailey et al., 2009), when it is unclear whether a couple is together or not. In both these cases, nonrelational labels such as FWBR may reflect (or gloss over) relational uncertainty or minimize perceptions of relational involvement to both the partner and the social network (Epstein et al., 2009). The FWBR label might also be useful in dealing with the social network. For example, when a romantic transition was attempted but failed, the FWBR label may be superior to admitting that a couple tried dating, but it did not pan out. In such cases, the FWBR label is used only retrospectively. The FWBR label can also hide relational uncertainty from the social network just as it can with the partner. Calling a potentially budding romantic relationship a FWBR might be superior to admitting uncertainty about the relational definition, trajectory, and the partner’s motivations. Finally, the FWBR label might provide a socially appropriate label for serial hookups with a stranger. Such communicative practices further blur the distinction between relational and nonrelational sex (e.g., Epstein et al., 2009). Ambiguous use of the FWB label can be disadvantageous, as it can obscure differences in partners’ desires and expectations—that is, although two people explicitly agree to the FWBR label and appear to be on the same page, they might actually think about the relationship in fundamentally different ways. The common FWBR script (Epstein et al., 2009) could aggravate such misunderstandings. For example, Partner A might desire a ‘‘booty call,’’ whereas Partner B wants to move toward a romantic relationship. In such cases, the person who wants ‘‘more’’ is probably at a distinct disadvantage (Sprecher, Schmeeckle, & Felmlee, 2006). In summary, Study 1 provides evidence of multiple FWBR types suggested in the literature (e.g., Bisson & Levine, 2009; Furman & Hand, 2006; Hughes et al., 2005; Mongeau et al., 2003). Not all participants, however, had experiences with FWBRs. Thus, the data might contain both scripted (e.g., definitions from participants without FWBR experiences) and experiential elements. Such a combination might make the data look more similar to the dominant script than if we had included only experiential data. Moreover, given that we did not initially set out to identify FWBR types, many questions remained. Therefore, Study 2 specifi- cally investigated differences among, and frequency of, the seven FWBR types. Study 2 Two primary questions drove Study 2. The most important goal was to validate the relational types. Specifically, FWBR types should differ systematically in nonsexual interactions, romantic history, and friendship strength. Second, Study 1 methods did not allow us to determine the frequency of the seven FWBR types. Therefore, we also considered how frequently each FWBR type occured (including gender differences in these reports). Validating FWBR Types Given our definitions, the seven FWBR types should differ in friendship strength, nonsexual interaction, and romantic history. To validate our typology, Study 2 focused on differences among FWBR types in these variables. Small and=or nonsignificant differences across FWBR types would greatly hinder the validity and utility of our typology. First, by definition, FWBRs types should differ in the frequency of nonsexual interactions and friendship strength. Specifically, FWBR types characterized by closeness and trust (e.g., true friends and successful transition in) should reflect more nonsexual interactions, and a stronger friendship at the initiation of sexual activity, than marginally-related partners or those who have contrasting relational goals (i.e., just sex and failed transition in). Other FWB types (network opportunism and transition out) were expected to fall between these extremes. Second, FWBR types should, by definition, differ in romantic experience. Participants in successful and unintentional transition in categories should report having a romantic relationship following their FWBR, whereas transition out partners should report a romantic relationship before the FWBR. Network opportunism, failed transition in, and (particularly) just sex FWBRs should be unlikely to include romantic experiences. VARIATION AMONG FRIENDS WITH BENEFITS 41
MONGEAU,KNIGHT.WILLIAMS.EDEN.AND SHAW Frequency of FWBR Types transition in.unintentional transition in. Given the dominant FWBR definition (e.g.,Epstein transition out,and network opportunism. H4b:Frequency of FWBR types will differ by gender. et al.,2009),one would expect the true friends type to Men will report being in just sex FWBRs more dominate students'reports.On the other hand,Epstein frequently than will women.Women will report et al.found that relatively few hookups follow the nor- being in transition in(failed,intentional,or unin- mative sexual script.Extending this thinking to FWBRs, tentional)and transition out FWBRs more than it is important to consider the frequency of FWB types men. because it will provide important evidence of the exist- ence of this relationship type.It is one thing to know that there are different types of FWBRs,but quite another to know how frequently these types Method appear.How many FWBRs follow the typical script Participants and procedures.Participants included (i.e.,true friends),how many deviate from it and in what 258 undergraduates(99 men,155 women,and four who ways? did not report;Mage=19.90,SD=1.93,range 18-30) Strong gender differences exist in early relational sex enrolled in upper-and lower-division communication (e.g.,Oliver Hyde,1993;Petersen Hyde,2010; classes at a large Southwestern U.S.university.Parti- Wells Twenge,2005).Kim et al.(2007)claimed that cipants predominately self-identified as heterosexual the dominant heterosexual sexual script depicts boys/ (92%)and White (nearly 70%).The sample overrepre- men as eschewing commitment and monogamy and sented freshmen (33.3%)and sophomores (34.1%),and going to great lengths to convince women to engage in underrepresented seniors (8.5%).Nearly 25%of parti- casual sex.Girls/women,on the other hand,prefer to cipants reported fraternity or sorority membership. pair sex with commitment and monogamy,and priori- Participants received extra credit for completing an tize relationships (Kim et al.,2007;see also Epstein online survey.To qualify,students must have had et al.,2009).In addition,"females may also use short- experience with a FWBR.Students without FWBR term relationships as a means of evaluating a male's experience could complete a different survey for equal suitability for a long-term relationship or securing his extra credit. interest so he will commit to a long-term relationship" (Cunningham Barbee,2008,p.99).Moreover, Measures women,when compared with men,tend to have more relational motivations for both hookups (Grello.Welsh. FWBR type.Participants read paragraphs describ- Harper,2006)and FWBRs (Lehmiller et al.,2010). ing seven FWBR types,and were asked to indicate the Therefore,we expect that men would be more likely to type that most closely matched their FWBR (contact report FWBRs as occurring independently from roman- the first author for a copy of the survey).An "other" tic relationships(e.g.,just sex),whereas women are more option was also provided. likely to link FWBRs to romantic relationships (e.g., transition in and transition out types). Romantic history.A single item probed if parti- Given our review,we posed the following hypotheses: cipants had a romantic relationship with their FWB partner.Response options were no;yes before the HI:Participants in successful transition in and true FWBR;yes after the FWBR;and yes,both before and friends FWBRs will report engaging in a wider after the FwBR. variety of nonsexual interaction than just sex and failed transition in types.Remaining FWBR types will fall between these extremes. Pre-sexual friendship.A single-item measure (i.e., H2:Participants in true friends and successful tran- "Before we had sex,our friendship was strong")tapped sition in will report stronger friendships,and just the strength of the pre-sexual friendship.The item was sex and failed transition in will report weaker accompanied by a five-point Likert scale ranging from friendships at the point of the first sexual interac- 1 (strongly disagree)to 5(strongly agree). tion than will other FWBR types. H3:Participants in successful and unintentional tran- sition in types will be more likely to report having Nonsexual activities.Six items tapped the extent to been in a romantic relationship following a which partners engaged in nonsexual activities (e.g., FWBR than the other types.Participants in tran- “We did lots of activities together,.”and“All we did sition out will be more likely to report having was have sex").In a principal components analysis with been in a romantic relationship before a FWBR. varimax rotation,four items factored together (=.90). Participants in remaining FWBR types will be The two items focusing exclusively engaging in sexual unlikely to describe any romantic history activity did not factor together or with the other four H4a:Reports of true friends will be more frequent than items.Given the importance of"All we did was have those for just sex,successful transition in.failed sex"ratings,we retained it as a single-item measure. 42
Frequency of FWBR Types Given the dominant FWBR definition (e.g., Epstein et al., 2009), one would expect the true friends type to dominate students’ reports. On the other hand, Epstein et al. found that relatively few hookups follow the normative sexual script. Extending this thinking to FWBRs, it is important to consider the frequency of FWB types because it will provide important evidence of the existence of this relationship type. It is one thing to know that there are different types of FWBRs, but quite another to know how frequently these types appear. How many FWBRs follow the typical script (i.e., true friends), how many deviate from it and in what ways? Strong gender differences exist in early relational sex (e.g., Oliver & Hyde, 1993; Petersen & Hyde, 2010; Wells & Twenge, 2005). Kim et al. (2007) claimed that the dominant heterosexual sexual script depicts boys= men as eschewing commitment and monogamy and going to great lengths to convince women to engage in casual sex. Girls=women, on the other hand, prefer to pair sex with commitment and monogamy, and prioritize relationships (Kim et al., 2007; see also Epstein et al., 2009). In addition, ‘‘females may also use shortterm relationships as a means of evaluating a male’s suitability for a long-term relationship or securing his interest so he will commit to a long-term relationship’’ (Cunningham & Barbee, 2008, p. 99). Moreover, women, when compared with men, tend to have more relational motivations for both hookups (Grello, Welsh, & Harper, 2006) and FWBRs (Lehmiller et al., 2010). Therefore, we expect that men would be more likely to report FWBRs as occurring independently from romantic relationships (e.g., just sex), whereas women are more likely to link FWBRs to romantic relationships (e.g., transition in and transition out types). Given our review, we posed the following hypotheses: H1: Participants in successful transition in and true friends FWBRs will report engaging in a wider variety of nonsexual interaction than just sex and failed transition in types. Remaining FWBR types will fall between these extremes. H2: Participants in true friends and successful transition in will report stronger friendships, and just sex and failed transition in will report weaker friendships at the point of the first sexual interaction than will other FWBR types. H3: Participants in successful and unintentional transition in types will be more likely to report having been in a romantic relationship following a FWBR than the other types. Participants in transition out will be more likely to report having been in a romantic relationship before a FWBR. Participants in remaining FWBR types will be unlikely to describe any romantic history. H4a: Reports of true friends will be more frequent than those for just sex, successful transition in, failed transition in, unintentional transition in, transition out, and network opportunism. H4b: Frequency of FWBR types will differ by gender. Men will report being in just sex FWBRs more frequently than will women. Women will report being in transition in (failed, intentional, or unintentional) and transition out FWBRs more than men. Method Participants and procedures. Participants included 258 undergraduates (99 men, 155 women, and four who did not report; Mage ¼ 19.90, SD ¼ 1.93, range ¼ 18–30) enrolled in upper- and lower-division communication classes at a large Southwestern U.S. university. Participants predominately self-identified as heterosexual (92%) and White (nearly 70%). The sample overrepresented freshmen (33.3%) and sophomores (34.1%), and underrepresented seniors (8.5%). Nearly 25% of participants reported fraternity or sorority membership. Participants received extra credit for completing an online survey. To qualify, students must have had experience with a FWBR. Students without FWBR experience could complete a different survey for equal extra credit. Measures FWBR type. Participants read paragraphs describing seven FWBR types, and were asked to indicate the type that most closely matched their FWBR (contact the first author for a copy of the survey). An ‘‘other’’ option was also provided. Romantic history. A single item probed if participants had a romantic relationship with their FWB partner. Response options were no; yes before the FWBR; yes after the FWBR; and yes, both before and after the FWBR. Pre-sexual friendship. A single-item measure (i.e., ‘‘Before we had sex, our friendship was strong’’) tapped the strength of the pre-sexual friendship. The item was accompanied by a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Nonsexual activities. Six items tapped the extent to which partners engaged in nonsexual activities (e.g., ‘‘We did lots of activities together,’’ and ‘‘All we did was have sex’’). In a principal components analysis with varimax rotation, four items factored together (a ¼ .90). The two items focusing exclusively engaging in sexual activity did not factor together or with the other four items. Given the importance of ‘‘All we did was have sex’’ ratings, we retained it as a single-item measure. MONGEAU, KNIGHT, WILLIAMS, EDEN, AND SHAW 42
VARIATION AMONG FRIENDS WITH BENEFITS Results Table 2. Romantic History across FWBR Types Validating FWBR types.HI predicted FWBR type Romantic History differences in the amount of nonsexual interaction.Con- sistent with this prediction,results of a one-way analysis Before After Both Before and FWBR Type None FWBR FWBR After FWBR of variance (ANOVA)indicated that mean values for nonsexual activities differed substantially across FWBR True friends 44[66.7%13[19.7% 5[7.6% 4[6.1% types,F6,214)=12.13,p<.001(=.25;see Table1). Just sex 3093.8% 1[3.1% 1[3.1% 00.0% Participants who indicated true friends and uninten- Network 28[75.7% 4[10.8%] 4[10.8a 12.7% opportunism tional transition in reflected the greatest nonsexual inter- Successful 7[33.3%2[9.5% 10[47.0% 29.5% action,whereas those indicating network opportunism transition in and(particularly)just sex reported the least. Unintentional 6[31.6%2「10.5%1947.5%1 2[10.5% Only had sex ratings (which did not factor with non- transition in sexual interaction)are also relevant to HI.Results from Failed 14[82.41 3[17.6%0[0.0% 0[0.0% transition in a one-way ANOVA indicated that mean ratings of Transition out 414.3%13[46.4%]310.7% 828.6% only had sex also differed across FWBR types,F(6, Total 133[60.5%38[17.3%32[14.5% 17[7.7% 214)=6.28,p<.001 (n=.15;see Table 1).In this case, participants who indicated failed transition in,transition Note.Within-row percentages are in brackets.FWBR=friends with benefits relationships. out,and just sex types reported significantly higher scores than those who indicated network opportunism, true friends.unintentional transition in.and successful out FWBRs were romantic both before and after the transition in types. FWBR.Failed transition in,network opportunism, H2 predicted that friendship strength at first sex and just sex FWBRs were unlikely to be romantic. would significantly differ across FWBR types.Overall, first sex friendship strength was moderate (M=4.57). Frequency of FWBR types.H4a focused on the Moreover,consistent with our prediction,mean values extent to which reports of true friends would be more for friendship strength dramatically varied across frequent than the other FWBR types.The number and FWBR types,,F6,214)=25.69,p<.001(2=.22;see percentage of participants indicating each FWBR type Table 1).Participants who indicated true friends and are provided in the far right-hand column of Table 3. transition out types reported the strongest friendships, True friends was the most frequent single type indicated, whereas participants indicating unintentional transition but only accounted for one-fourth of responses.True in and just sex reported the weakest friendships. friends were followed(in descending order of frequency) H3 predicted differences in the extent and timing by network opportunism,just sex,other,transition out of romantic relationship history across FWBR types. and successful,unintentional,and failed transition in. Overall,39.5%of participants reported a romantic rela- H4b predicted differences in the frequency with tionship either before or after the FWBR (or both). which men and women would indicate FWBR types. Consistent with H3,romantic history responses Specifically,men were expected to be more likely than substantially differed across FWBR types,(18,N= women to report just sex,and women more likely than 220)=97.95,p<.001(Cramer'sV=.39;see Table2). men to report transition in and transition out FWBRs. Successful and unintentional transition in types were Given that 60%of the sample was women,if gender is much more likely to be romantic following,and tran- unrelated to frequency,all seven types should exhibit sition out before,an FWBR.Nearly 30%of transition Table 3.Frequencies of FWBR Types.Both Overall and Table 1.Mean Values of Nonsexual Activities,Only Had Sex. Broken Down by Participant Gender and Friendship Strength Broken Down by FWBR Type Participant Gender Nonsexual Only Had Friendship FWBR Type Men Women Total Frequency FWBR Type Activities Sex Strength True friends 23 42 65[26.1% True friends 5.40 3.18p 5.39 Just sex 20 12 32[12.4% Just sex 3.12 5.09. 2.91d Network opportunism 11 25 36[14.5% Network opportunism 4.1b 3.27b 4.49b.c Successful transition in 10 21[8.4% Successful transition in 4.76b 2.90. 4.71a Unintentional transition in 6 13 19[7.6%] Unintentional transition in 5.16c 2.95b 3.89e Failed transition in 2 15 17[6.8% Failed transition in 4.59b.c 4.33 4.33b. Transition out 15 13 2811.2% Transition out 4.71b.c 4.29 5.18.b Other 10 21 3112.4%] Overall mean 4.62 3.66 4.57 Total 97 152 249[100.0% Note.Within columns,means lacking a common subscript differed Note.Within-column percentages are in brackets. FWBR=friends significantly (p<.05).FWBR=friends with benefits relationships. with benefits relationships. 43
Results Validating FWBR types. H1 predicted FWBR type differences in the amount of nonsexual interaction. Consistent with this prediction, results of a one-way analysis of variance (ANOVA) indicated that mean values for nonsexual activities differed substantially across FWBR types, F(6, 214) ¼ 12.13, p < .001 (g2 p ¼ :25; see Table 1). Participants who indicated true friends and unintentional transition in reflected the greatest nonsexual interaction, whereas those indicating network opportunism and (particularly) just sex reported the least. Only had sex ratings (which did not factor with nonsexual interaction) are also relevant to H1. Results from a one-way ANOVA indicated that mean ratings of only had sex also differed across FWBR types, F(6, 214) ¼ 6.28, p < .001 (g2 p ¼ :15; see Table 1). In this case, participants who indicated failed transition in, transition out, and just sex types reported significantly higher scores than those who indicated network opportunism, true friends, unintentional transition in, and successful transition in types. H2 predicted that friendship strength at first sex would significantly differ across FWBR types. Overall, first sex friendship strength was moderate (M ¼ 4.57). Moreover, consistent with our prediction, mean values for friendship strength dramatically varied across FWBR types, F(6, 214) ¼ 25.69, p < .001 (g2 p ¼ :22; see Table 1). Participants who indicated true friends and transition out types reported the strongest friendships, whereas participants indicating unintentional transition in and just sex reported the weakest friendships. H3 predicted differences in the extent and timing of romantic relationship history across FWBR types. Overall, 39.5% of participants reported a romantic relationship either before or after the FWBR (or both). Consistent with H3, romantic history responses substantially differed across FWBR types, v2 (18, N ¼ 220) ¼ 97.95, p < .001 (Cramer’s V ¼ .39; see Table 2). Successful and unintentional transition in types were much more likely to be romantic following, and transition out before, an FWBR. Nearly 30% of transition out FWBRs were romantic both before and after the FWBR. Failed transition in, network opportunism, and just sex FWBRs were unlikely to be romantic. Frequency of FWBR types. H4a focused on the extent to which reports of true friends would be more frequent than the other FWBR types. The number and percentage of participants indicating each FWBR type are provided in the far right-hand column of Table 3. True friends was the most frequent single type indicated, but only accounted for one-fourth of responses. True friends were followed (in descending order of frequency) by network opportunism, just sex, other, transition out and successful, unintentional, and failed transition in. H4b predicted differences in the frequency with which men and women would indicate FWBR types. Specifically, men were expected to be more likely than women to report just sex, and women more likely than men to report transition in and transition out FWBRs. Given that 60% of the sample was women, if gender is unrelated to frequency, all seven types should exhibit Table 1. Mean Values of Nonsexual Activities, Only Had Sex, and Friendship Strength Broken Down by FWBR Type FWBR Type Nonsexual Activities Only Had Sex Friendship Strength True friends 5.40a 3.18b 5.39b Just sex 3.12d 5.09a 2.91d Network opportunism 4.11b 3.27b 4.49b,c Successful transition in 4.76a,b 2.90b 4.71a,c Unintentional transition in 5.16a,c 2.95b 3.89c Failed transition in 4.59b,c 4.33a 4.33b,c Transition out 4.71b,c 4.29a 5.18a,b Overall mean 4.62 3.66 4.57 Note. Within columns, means lacking a common subscript differed significantly (p < .05). FWBR ¼ friends with benefits relationships. Table 2. Romantic History across FWBR Types Romantic History FWBR Type None Before FWBR After FWBR Both Before and After FWBR True friends 44 [66.7%] 13 [19.7%] 5 [7.6%] 4 [6.1%] Just sex 30 [93.8%] 1 [3.1%] 1 [3.1%] 0 [0.0%] Network opportunism 28 [75.7%] 4 [10.8%] 4 [10.8%] 1 [2.7%] Successful transition in 7 [33.3%] 2 [9.5%] 10 [47.0%] 2 [9.5%] Unintentional transition in 6 [31.6%] 2 [10.5%] 9 [47.5%] 2 [10.5%] Failed transition in 14 [82.4] 3 [17.6%] 0 [0.0%] 0 [0.0%] Transition out 4 [14.3%] 13 [46.4%] 3 [10.7%] 8 [28.6%] Total 133 [60.5%] 38 [17.3%] 32 [14.5%] 17 [7.7%] Note. Within-row percentages are in brackets. FWBR ¼ friends with benefits relationships. Table 3. Frequencies of FWBR Types, Both Overall and Broken Down by Participant Gender Participant Gender FWBR Type Men Women Total Frequency True friends 23 42 65 [26.1%] Just sex 20 12 32 [12.4%] Network opportunism 11 25 36 [14.5%] Successful transition in 10 11 21 [8.4%] Unintentional transition in 6 13 19 [7.6%] Failed transition in 2 15 17 [6.8%] Transition out 15 13 28 [11.2%] Other 10 21 31 [12.4%] Total 97 152 249 [100.0%] Note. Within-column percentages are in brackets. FWBR ¼ friends with benefits relationships. VARIATION AMONG FRIENDS WITH BENEFITS 43
MONGEAU,KNIGHT.WILLIAMS.EDEN.AND SHAW the 60-40 pattern (within sampling error).Analyses The three transition in FWBRs(successful,uninten- indicated proportions of men and women in the seven tional,and failed)differed in important and interesting FWBR types significantly differed,x2(7,N=249)= ways.Both successful and unintentional transition in 18.36,p<.01 (Cramer's V=.27;see Table 3).Men types were similar to true friends on reports of nonsex- represented a strong majority of the just sex FWBRs, ual interaction and only engaging in sex.Unintentional and women nearly all the failed transition in FWBRs. transition in cases,however,were lower in first sex Surprisingly,men were more likely than women to friendship strength than were successful transition in. report transition out FWBRs.Sex differences for true This suggests that unintentional transition in types friends,successful transition in,failed transition in,and began as hookups that blossomed,over time,into network opportunism matched the sample characteris- romantic relationships.Failed transition in cases,on tics.Therefore,these data partially support H4b. the other hand,were quite high on only had sex and near the overall average in nonsexual activities and friendship strength at the first sex. Study 2 Discussion Finally,the transition out type was relatively high in Sexual interaction on modern college campuses only had sex.but above average on friendship strength. centers on the hookup,where partners,neither involved Previous romantic history might make an ex-partner a nor interested in future interaction,engage in sexual particularly attractive sexual target,as sexual interac- interaction (Bogle,2008;Paul Hayes,2002).Along tion may represent a strategic attempt to rekindle the similar lines,FWBRs occur when friends have sex on romantic flame.Consistent with this notion,nearly repeated occasions without the expectations of a roman- one-third of transition out relationships were romantic tic transition (Bisson Levine,2009;Epstein et al., both before and after the FWBR (see Dailey et al.. 2009;Hughes et al.,2005).Data from Study 1 led us 2009).In these cases,FWBR might be a simple label to reject the common assumption that FWBRs represent that helps partners evaluate,communicate,and poten- a single relationship type,and suggested seven FWBRs tially maintain the ambiguous and dynamic nature of types (i.e.,true friends,just sex,network opportunism, their relationship. transition out,as well as successful,unintentional,and This is not to say,however,that our results are uni- failed transition in).Study 2 was designed to validate versally consistent with our expectations.For example, the FWBR types and to identify how frequently these approximately one-third of both unintentional and suc- various types occur. cessful transition in participants reported that they never had a romantic relationship with their partner.(To a les- ser extent,the same pattern appears for transition out as Validating FWBR types.Our conceptualizations of well.)Perhaps participants were willing to indicate that the various FWBR types assume differences in non- their FWBR most closely approximated a transition sexual interaction,friendship strength,and romantic in:however,when asked directly.,they may have been history.As a consequence,failure to find robust differ- unwilling to characterize their relationship as romantic. ences on these variables across FWBR types would This suggests investigating college students'meanings greatly damage the validity of our typology.Fortu- for both romantic and casual relationships would be nately,all of the differences among FWBR types were insightful. statistically significant;moderate to large in size (Cohen,1988);and,for the most part,consistent with our hypotheses Frequency of FWBR types.Following Epstein True friends and just sex types anchor the extremes of et al.'s(2009)discussion of hookups,we assumed that, the FWBR types.Partners in true friends-the proto- although college students describe FWBRs similarly, typical FWBR (Epstein et al.,2009;Hughes et al., not all cases will follow that form.Participants 2005)-reported engaging in the most nonsexual activi- reported true friends (most consistent with the FWBR ties,were least likely to report only having sex,and definition)most frequently;however,they represented reported very high friendship strength at the point of only one-fourth of cases.Thus,in Study 2 data,most first sex.Just sex FWBRs.on the other hand,were asso- cases diverged from the dominant FWBR definition. ciated with (by far)the least nonsexual interaction, For example,over one-third of FWBRs involved a weakest first sex friendship,and the most exclusive romantic relationship (either before or after the sexual interaction.Network opportunism FWBRs fell FWBR).and one in eight FWBRs involved partners between these extremes,as they reflected relatively low only having sex. ratings on both nonsexual interaction and only having Men and women also reported the FWBR types in sex.This is consistent with Afifi and Faulkner's (2000) significantly different proportions.Based on the pre- notion that sexual activity represents another form of dominant sexual script (Kim et al.,2007),we expected socializing among heterosexual college students (like men to report FWBRs that involved few relationship self-disclosure or engaging in drinking games). entanglements (i.e.,just sex),whereas women would be 44
the 60–40 pattern (within sampling error). Analyses indicated proportions of men and women in the seven FWBR types significantly differed, v2 (7, N ¼ 249) ¼ 18.36, p < .01 (Cramer’s V ¼ .27; see Table 3). Men represented a strong majority of the just sex FWBRs, and women nearly all the failed transition in FWBRs. Surprisingly, men were more likely than women to report transition out FWBRs. Sex differences for true friends, successful transition in, failed transition in, and network opportunism matched the sample characteristics. Therefore, these data partially support H4b. Study 2 Discussion Sexual interaction on modern college campuses centers on the hookup, where partners, neither involved nor interested in future interaction, engage in sexual interaction (Bogle, 2008; Paul & Hayes, 2002). Along similar lines, FWBRs occur when friends have sex on repeated occasions without the expectations of a romantic transition (Bisson & Levine, 2009; Epstein et al., 2009; Hughes et al., 2005). Data from Study 1 led us to reject the common assumption that FWBRs represent a single relationship type, and suggested seven FWBRs types (i.e., true friends, just sex, network opportunism, transition out, as well as successful, unintentional, and failed transition in). Study 2 was designed to validate the FWBR types and to identify how frequently these various types occur. Validating FWBR types. Our conceptualizations of the various FWBR types assume differences in nonsexual interaction, friendship strength, and romantic history. As a consequence, failure to find robust differences on these variables across FWBR types would greatly damage the validity of our typology. Fortunately, all of the differences among FWBR types were statistically significant; moderate to large in size (Cohen, 1988); and, for the most part, consistent with our hypotheses. True friends and just sex types anchor the extremes of the FWBR types. Partners in true friends—the prototypical FWBR (Epstein et al., 2009; Hughes et al., 2005)—reported engaging in the most nonsexual activities, were least likely to report only having sex, and reported very high friendship strength at the point of first sex. Just sex FWBRs, on the other hand, were associated with (by far) the least nonsexual interaction, weakest first sex friendship, and the most exclusive sexual interaction. Network opportunism FWBRs fell between these extremes, as they reflected relatively low ratings on both nonsexual interaction and only having sex. This is consistent with Afifi and Faulkner’s (2000) notion that sexual activity represents another form of socializing among heterosexual college students (like self-disclosure or engaging in drinking games). The three transition in FWBRs (successful, unintentional, and failed) differed in important and interesting ways. Both successful and unintentional transition in types were similar to true friends on reports of nonsexual interaction and only engaging in sex. Unintentional transition in cases, however, were lower in first sex friendship strength than were successful transition in. This suggests that unintentional transition in types began as hookups that blossomed, over time, into romantic relationships. Failed transition in cases, on the other hand, were quite high on only had sex and near the overall average in nonsexual activities and friendship strength at the first sex. Finally, the transition out type was relatively high in only had sex, but above average on friendship strength. Previous romantic history might make an ex-partner a particularly attractive sexual target, as sexual interaction may represent a strategic attempt to rekindle the romantic flame. Consistent with this notion, nearly one-third of transition out relationships were romantic both before and after the FWBR (see Dailey et al., 2009). In these cases, FWBR might be a simple label that helps partners evaluate, communicate, and potentially maintain the ambiguous and dynamic nature of their relationship. This is not to say, however, that our results are universally consistent with our expectations. For example, approximately one-third of both unintentional and successful transition in participants reported that they never had a romantic relationship with their partner. (To a lesser extent, the same pattern appears for transition out as well.) Perhaps participants were willing to indicate that their FWBR most closely approximated a transition in; however, when asked directly, they may have been unwilling to characterize their relationship as romantic. This suggests investigating college students’ meanings for both romantic and casual relationships would be insightful. Frequency of FWBR types. Following Epstein et al.’s (2009) discussion of hookups, we assumed that, although college students describe FWBRs similarly, not all cases will follow that form. Participants reported true friends (most consistent with the FWBR definition) most frequently; however, they represented only one-fourth of cases. Thus, in Study 2 data, most cases diverged from the dominant FWBR definition. For example, over one-third of FWBRs involved a romantic relationship (either before or after the FWBR), and one in eight FWBRs involved partners only having sex. Men and women also reported the FWBR types in significantly different proportions. Based on the predominant sexual script (Kim et al., 2007), we expected men to report FWBRs that involved few relationship entanglements (i.e., just sex), whereas women would be MONGEAU, KNIGHT, WILLIAMS, EDEN, AND SHAW 44
VARIATION AMONG FRIENDS WITH BENEFITS more likely to link FWBRs to romantic relationships FWBR types also differed in relationship characteristics. (i.e.,transition in and transition out).Frequencies of These differences challenge notions of place,meaning, FWBR types for men and women were at the same time and enactment of heterosexual interaction on college consistent and inconsistent with the predominant sexual campuses. script of casual sex.Men were indeed more likely to report just sex FWBRs;however,they reported the majority of transition out FWBRs (despite representing Relational Implications of FWBR Types only 39%of the sample).Women substantially exceeded The common FWBRs script describes sexual interac- men only in the failed transition in type. tion between good friends who eschew the drama of Given our thinking,there are two primary explana- romantic relationships (e.g.,Bisson Levine,2009; tions for these gender differences.The first potential Epstein et al.,2009).Our data,however,suggest sub- explanation focuses on the way that FWBRs are labeled. stantial variation in relational histories and motivations For example,women might be more willing to label across FWBRs.Successful transition in FWBRs rep- failed transition in FWBRs to put a positive spin on a resent an intentional attempt at a romantic relationship relational failure.Given the sexual double standard,a transition,whereas unintentional FWBRs represent the woman might be negatively evaluated for trying to byproduct of sexual interaction and associated interac- initiate a romantic relationship with a man through tions.We suspect that the same is likely true of tran- sexual activity.Calling the episode a FWBR might pro- sition out FWBRs,where nearly 30%of cases are vide a more relational explanation that would not be so romantic both before and after the FWBR,particularly harshly evaluated. among men who want to hang onto a faltering An alternative explanation for differences is that men relationship. and women strategically use FWBRs in different ways. Romantic attraction and motivations,(i.e.,a desire for Women might report nearly all the failed transition in something more)represents the raison d'etre for some FWBRs because,as Cunningham and Barbee(2008)sug- FWBRs.Thus,some FWBRs(e.g.,successful transition gested,they use FWBRs to test the man's suitability and in and some transition out)have a betweenness quality. interest in committing to a romantic relationship.In such By betweenness,we mean that some FWBRs contain ele- a case,a man's lack of interest in commitment would ments of both friendships and romantic relationships, lead to a failed transition in.Moreover,men are more and are strategically designed to initiate romantic rela- likely to report transition out FWBRs than are women. tionships.The notion of betweenness likely adds a level This may be relevant to Rubin,Peplau,and Hill's of uncertainty and complexity to FWBR experiences (1981)contention that men are less willing to break off because,although they are not supposed to be romantic relationships than are women.Men,in this case,might or involve romantic motivations,some clearly are.The be willing to use FWBRs as a means of forestalling per- transition from a FWBR to a romantic relationship, manent romantic termination by hanging onto and moreover,is unlikely to be simple.Many FWBRs are dif- potentially repairing a floundering relationship. ficult to maintain because romantic attraction and inter- Given these data,we cannot differentiate between est is frequently unilateral (Bisson Levine,2009; labeling and actual relational explanations for our gen- Hughes et al.,2005:Mongeau et al..2003).In such cases. der differences.Labeling differences are certainly poss- FWBRs may represent a relational compromise (Bisson ible even within the same FWBRs.As we have argued, Levine,2009;Mongeau et al.,2003),particularly for partners might consider their FWBRs as different types women(Cunningham Barbee,2008). (or one may consider the entanglement an FWBR, Even when romantic desire is mutual.the transition whereas the other may not).Labels will also likely from a FWBR to a romantic entanglement is likely change across the course of an FWBR (e.g.,the tran- difficult.In the 1970s,the predominant sexual script sition in cases)such that labeling itself,and agreement prescribed that sexual intimacy should match psycholog- on a label,might be a very complex enterprise. ical and communicative intimacy (Perlman Sprecher,in press;Sprecher,1989).In most FWBRs,however,part- ners engage in sexual activity before the acknowledgment General Discussion of romantic attraction and attachment.Metts (2004) reported that engaging in significant sexual activity before Counter to the general claim that FWBRs represent a saying"I love you"is negatively related to relationship singular relationship type (e.g.,Bisson Levine,2009; development and positively related to regret following Hughes et al.,2005),these studies identified substantial the sexual interaction. variation in FWBRs.Consistent with the notion that There are several implications stemming from the nonrelational sex practices vary from shared definitions potentials and pitfalls of the FWBR to romantic relation- (Epstein et al.,2009),Study 1 identified seven types of ship transition.First,this particular transition does not FWBRs.In Study 2,nearly 90%of participants selected happen very often.Only about 15%of FWBRs success- one of these types in a forced-choice measure.The seven fully make the transition to romantic relationships,and 45
more likely to link FWBRs to romantic relationships (i.e., transition in and transition out). Frequencies of FWBR types for men and women were at the same time consistent and inconsistent with the predominant sexual script of casual sex. Men were indeed more likely to report just sex FWBRs; however, they reported the majority of transition out FWBRs (despite representing only 39% of the sample). Women substantially exceeded men only in the failed transition in type. Given our thinking, there are two primary explanations for these gender differences. The first potential explanation focuses on the way that FWBRs are labeled. For example, women might be more willing to label failed transition in FWBRs to put a positive spin on a relational failure. Given the sexual double standard, a woman might be negatively evaluated for trying to initiate a romantic relationship with a man through sexual activity. Calling the episode a FWBR might provide a more relational explanation that would not be so harshly evaluated. An alternative explanation for differences is that men and women strategically use FWBRs in different ways. Women might report nearly all the failed transition in FWBRs because, as Cunningham and Barbee (2008) suggested, they use FWBRs to test the man’s suitability and interest in committing to a romantic relationship. In such a case, a man’s lack of interest in commitment would lead to a failed transition in. Moreover, men are more likely to report transition out FWBRs than are women. This may be relevant to Rubin, Peplau, and Hill’s (1981) contention that men are less willing to break off relationships than are women. Men, in this case, might be willing to use FWBRs as a means of forestalling permanent romantic termination by hanging onto and potentially repairing a floundering relationship. Given these data, we cannot differentiate between labeling and actual relational explanations for our gender differences. Labeling differences are certainly possible even within the same FWBRs. As we have argued, partners might consider their FWBRs as different types (or one may consider the entanglement an FWBR, whereas the other may not). Labels will also likely change across the course of an FWBR (e.g., the transition in cases) such that labeling itself, and agreement on a label, might be a very complex enterprise. General Discussion Counter to the general claim that FWBRs represent a singular relationship type (e.g., Bisson & Levine, 2009; Hughes et al., 2005), these studies identified substantial variation in FWBRs. Consistent with the notion that nonrelational sex practices vary from shared definitions (Epstein et al., 2009), Study 1 identified seven types of FWBRs. In Study 2, nearly 90% of participants selected one of these types in a forced-choice measure. The seven FWBR types also differed in relationship characteristics. These differences challenge notions of place, meaning, and enactment of heterosexual interaction on college campuses. Relational Implications of FWBR Types The common FWBRs script describes sexual interaction between good friends who eschew the drama of romantic relationships (e.g., Bisson & Levine, 2009; Epstein et al., 2009). Our data, however, suggest substantial variation in relational histories and motivations across FWBRs. Successful transition in FWBRs represent an intentional attempt at a romantic relationship transition, whereas unintentional FWBRs represent the byproduct of sexual interaction and associated interactions. We suspect that the same is likely true of transition out FWBRs, where nearly 30% of cases are romantic both before and after the FWBR, particularly among men who want to hang onto a faltering relationship. Romantic attraction and motivations, (i.e., a desire for something more) represents the raison d’eˆtre for some FWBRs. Thus, some FWBRs (e.g., successful transition in and some transition out) have a betweenness quality. By betweenness, we mean that some FWBRs contain elements of both friendships and romantic relationships, and are strategically designed to initiate romantic relationships. The notion of betweenness likely adds a level of uncertainty and complexity to FWBR experiences because, although they are not supposed to be romantic or involve romantic motivations, some clearly are. The transition from a FWBR to a romantic relationship, moreover, is unlikely to be simple. Many FWBRs are dif- ficult to maintain because romantic attraction and interest is frequently unilateral (Bisson & Levine, 2009; Hughes et al., 2005; Mongeau et al., 2003). In such cases, FWBRs may represent a relational compromise (Bisson & Levine, 2009; Mongeau et al., 2003), particularly for women (Cunningham & Barbee, 2008). Even when romantic desire is mutual, the transition from a FWBR to a romantic entanglement is likely difficult. In the 1970s, the predominant sexual script prescribed that sexual intimacy should match psychological and communicative intimacy (Perlman & Sprecher, in press; Sprecher, 1989). In most FWBRs, however, partners engage in sexual activity before the acknowledgment of romantic attraction and attachment. Metts (2004) reported that engaging in significant sexual activity before saying ‘‘I love you’’ is negatively related to relationship development and positively related to regret following the sexual interaction. There are several implications stemming from the potentials and pitfalls of the FWBR to romantic relationship transition. First, this particular transition does not happen very often. Only about 15% of FWBRs successfully make the transition to romantic relationships, and VARIATION AMONG FRIENDS WITH BENEFITS 45
MONGEAU,KNIGHT,WILLIAMS,EDEN.AND SHAW only one-half of those were intentionally performed. Calabrese, 1975;Solomon Knobloch,2004; Although this is not a particularly large proportion Sunnafrank.1986)and sensemaking (Weick.2001)are (Bisson Levine,2009;Perlman Sprecher,in press), likely quite useful in extending future understanding data for other transitions (e.g.,from first dates)are not how FWBRs develop. available.It is possible that only about 15%of first dates end up generating a romantic relationship.In addition, our discussion suggests that the better partners know Limitations and Directions for Future Research one another(e.g.,true friends over network opportunism Our results should be considered in light of several over just sex),the more successful the FWBR to romantic important limitations.First,our samples were limited relationship transition might be.Thus,not all FWBRs to college students at two U.S.universities.It is unclear are the same in terms of their romantic potential.Cou- how these results would generalize to other U.S.univer- ples'past communicative and physical interactions likely sities (or those in other countries),participants'non- influence the nature.and sustainability.of a future college-age cohorts,older adults,or high school stu- romantic relationship dents.Second,that students recalled past behaviors makes the direction of causality ambiguous.Third,we Making Sense of FWBRs placed cases into FWBR types based on only one partner's report.The other partners'definitions and By definition,FWBRs defy traditional scripts for sex motivations might differ. (i.e.,sex reflects intimacy and closeness;Bogle,2008) The FWBR label covers multiple relationship types, and friendship (i.e.,friends do not have sex;Werking, suggesting that it is much more complex than previously 1997).In addition,the multiple types of FWBRs indi- suggested (e.g.,Bisson Levine,2009;Hughes et al., cate that they occur in a variety of social and relational 2005).Therefore,future FWBR research should include contexts.Some FWBRs have replaced dates as a vehicle longitudinal and dyadic data collection.Such methods. for investigating romantic potential,whereas others are however,are likely complicated because some partners serial hookups.Given this variety,how do we,as scho- are strangers at the first sexual interaction,making lars,make sense of FWBRs?Has the FWBR term been dyadic first sex data difficult to obtain.In other cases, stretched so far that the moniker has effectively lost its partners apply the FWBR label after sex begins(and, meaning?Our contention is that it has not.These data in some cases,after it ends).As such,there is likely to suggest that the term FWBR is considerably broader no such thing as an ideal FWBR study.A combination than the label (and initial scholarship)suggested.How- of longitudinal.recall.and interview studies will likely ever,the same is true of the study of hookups.Initially, most effectively triangulate on the complex and mal- hookups were defined as single sexual episodes between leable nature of FWBRs strangers(e.g.,Paul Hayes,2002);however,over time, definitions of the term expanded (e.g.,Esptein et al.. 2009)as scholars were able to adjust their definitions References and approaches to study the "new"version of the phenomenon.It is also possible that the very nature of Afifi,W.A..Faulkner,S.L.(2000).On being "just friends":The FWBRs has changed over time.When the phenomenon frequency and impact of sexual activity in cross-sex friendships. (or at least the label)appeared in the late 1990s,it may Journal of Social and Personal Relationships.17.205-222.doi: 10.1177/0265407500172003 have predominately appeared as true friends(as FWBR Bailey,B.L.(1988).From front porch to back seat:Courtship in suggests).Over time,however,the FWBR label might twentieth-century America.Baltimore.MD:Johns Hopkins have expanded to fill the cracks between other relational University Press. labels. Bailey,B.L.(1999).Sex in the heartland:Politics,culture,and the Making sense of FWBRs is likely difficult for sexual/ sexual revolution.Cambridge,MA:Harvard University Press. relational partners as well.First sex with a FWB partner Berger,C.R.,Calabrese,R.J.(1975).Some explorations in initial interaction and beyond:Toward a developmental theory of has many potential relational interpretations (e.g.,a interpersonal communication.Human Communication Research, hookup,caring for a close friend,or a romantic relation- 1,99-112.doi:10.1111/j.1468-2958.1975.tb00258.x ship transition)that likely take time to differentiate. Bisson,M.A.,Levine,T.R.(2009).Negotiating a friends with Whereas some instances are likely acknowledged as benefits relationship.Archives of Sexual Behavior,38,66-73 doi:10.1007/s10508-007-9211-2 FWBRs even before the onset of sexual activity,other Bleske,A.L.,Buss,D.M.(2000).Can men and women be just instances are only identified midstream,whereas still friends?Personal Relationships,21,131-151.doi:10.1111/ others are only labeled in retrospect.An important set j.1475-6811.2000.tb00008.x of questions,then,is how partners come to understand Bogle,K.A.(2008).Hooking up:Sex dating,and relationships on that they are in an FWBR,when they come to that campus.New York,NY:New York University Press. realization,and the information that they use to make Cohen,J.(1988).Statistical power analysis for the behavioral sciences (2nd ed.).Hillsdale,NJ:Lawrence Erlbaum Associates,Inc. that attribution.Therefore,we believe that theories Cunningham,M.R.,Barbee,A.P.(2008).Prelude to a kiss:Non- focused on uncertainty and their reduction (Berger verbal flirting,opening gambits,and other communication 46
only one-half of those were intentionally performed. Although this is not a particularly large proportion (Bisson & Levine, 2009; Perlman & Sprecher, in press), data for other transitions (e.g., from first dates) are not available. It is possible that only about 15% of first dates end up generating a romantic relationship. In addition, our discussion suggests that the better partners know one another (e.g., true friends over network opportunism over just sex), the more successful the FWBR to romantic relationship transition might be. Thus, not all FWBRs are the same in terms of their romantic potential. Couples’ past communicative and physical interactions likely influence the nature, and sustainability, of a future romantic relationship. Making Sense of FWBRs By definition, FWBRs defy traditional scripts for sex (i.e., sex reflects intimacy and closeness; Bogle, 2008) and friendship (i.e., friends do not have sex; Werking, 1997). In addition, the multiple types of FWBRs indicate that they occur in a variety of social and relational contexts. Some FWBRs have replaced dates as a vehicle for investigating romantic potential, whereas others are serial hookups. Given this variety, how do we, as scholars, make sense of FWBRs? Has the FWBR term been stretched so far that the moniker has effectively lost its meaning? Our contention is that it has not. These data suggest that the term FWBR is considerably broader than the label (and initial scholarship) suggested. However, the same is true of the study of hookups. Initially, hookups were defined as single sexual episodes between strangers (e.g., Paul & Hayes, 2002); however, over time, definitions of the term expanded (e.g., Esptein et al., 2009) as scholars were able to adjust their definitions and approaches to study the ‘‘new’’ version of the phenomenon. It is also possible that the very nature of FWBRs has changed over time. When the phenomenon (or at least the label) appeared in the late 1990s, it may have predominately appeared as true friends (as FWBR suggests). Over time, however, the FWBR label might have expanded to fill the cracks between other relational labels. Making sense of FWBRs is likely difficult for sexual= relational partners as well. First sex with a FWB partner has many potential relational interpretations (e.g., a hookup, caring for a close friend, or a romantic relationship transition) that likely take time to differentiate. Whereas some instances are likely acknowledged as FWBRs even before the onset of sexual activity, other instances are only identified midstream, whereas still others are only labeled in retrospect. An important set of questions, then, is how partners come to understand that they are in an FWBR, when they come to that realization, and the information that they use to make that attribution. Therefore, we believe that theories focused on uncertainty and their reduction (Berger & Calabrese, 1975; Solomon & Knobloch, 2004; Sunnafrank, 1986) and sensemaking (Weick, 2001) are likely quite useful in extending future understanding how FWBRs develop. Limitations and Directions for Future Research Our results should be considered in light of several important limitations. First, our samples were limited to college students at two U.S. universities. It is unclear how these results would generalize to other U.S. universities (or those in other countries), participants’ noncollege-age cohorts, older adults, or high school students. Second, that students recalled past behaviors makes the direction of causality ambiguous. Third, we placed cases into FWBR types based on only one partner’s report. The other partners’ definitions and motivations might differ. The FWBR label covers multiple relationship types, suggesting that it is much more complex than previously suggested (e.g., Bisson & Levine, 2009; Hughes et al., 2005). Therefore, future FWBR research should include longitudinal and dyadic data collection. Such methods, however, are likely complicated because some partners are strangers at the first sexual interaction, making dyadic first sex data difficult to obtain. In other cases, partners apply the FWBR label after sex begins (and, in some cases, after it ends). As such, there is likely to no such thing as an ideal FWBR study. A combination of longitudinal, recall, and interview studies will likely most effectively triangulate on the complex and malleable nature of FWBRs. References Afifi, W. A., & Faulkner, S. L. (2000). On being ‘‘just friends’’: The frequency and impact of sexual activity in cross-sex friendships. Journal of Social and Personal Relationships, 17, 205–222. doi: 10.1177=0265407500172003 Bailey, B. L. (1988). From front porch to back seat: Courtship in twentieth-century America. Baltimore, MD: Johns Hopkins University Press. Bailey, B. L. (1999). Sex in the heartland: Politics, culture, and the sexual revolution. Cambridge, MA: Harvard University Press. Berger, C. R., & Calabrese, R. J. (1975). Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1, 99–112. doi: 10.1111=j.1468-2958.1975.tb00258.x Bisson, M. A., & Levine, T. R. (2009). Negotiating a friends with benefits relationship. Archives of Sexual Behavior, 38, 66–73. doi: 10.1007=s10508-007-9211-2 Bleske, A. L., & Buss, D. M. (2000). Can men and women be just friends? Personal Relationships, 21, 131–151. doi: 10.1111= j.1475-6811.2000.tb00008.x Bogle, K. A. (2008). Hooking up: Sex dating, and relationships on campus. New York, NY: New York University Press. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Cunningham, M. R., & Barbee, A. P. (2008). Prelude to a kiss: Nonverbal flirting, opening gambits, and other communication MONGEAU, KNIGHT, WILLIAMS, EDEN, AND SHAW 46