The Coming-Out Process of Young Lesbian and Bisexual Women: Are There Butch/Femme Differences in Sexual Identity Development?

Abstract

Research on lesbian and bisexual women has documented various biological and behavioral differences between butch and femme women. However, little research has examined whether differences exist in sexual identity development (i.e., the coming-out process). The present study examined longitudinally potential butch/femme differences in sexual identity formation and integration among an ethnically diverse sample of 76 self-identified lesbian and bisexual young women (ages 14–21 years). A composite measure of butch/femme identity classified 43% as butch and 51% as femme. Initial comparisons found butch/femme differences in sexual identity (i.e., nearly all butches identified as lesbian, but about half of femmes identified as bisexual), suggesting the need to examine this confound. Comparisons of lesbian butches, lesbian femmes, and bisexual femmes found that lesbian butches and femmes generally did not differ on sexual identity formation, but they differed from bisexual femmes. Lesbian butches and femmes had sexual behaviors and a cognitive sexual orientation that were more centered on women than those of bisexual femmes. With respect to sexual identity integration, lesbian butches were involved in more gay social activities, were more comfortable with others knowing about their homosexuality, and were more certain, comfortable, and accepting of their sexual identity than were bisexual femmes. Fewer differences were found between lesbian femmes and bisexual femmes or between lesbian butches and lesbian femmes. The findings suggest that sexual identity formation does not differ between butch or femme women, but differences are linked to sexual identity as lesbian or bisexual. Further, the findings that lesbian femmes sometimes differed from lesbian butches and at other times from bisexual femmes on sexual identity integration suggest that neither sexual identity nor butch/femme alone may explain sexual identity integration. Research examining the intersection between sexual identity and butch/femme is needed.

INTRODUCTION

The sexual identity development of lesbian, gay, and bisexual (LGB) individuals, also known as the coming-out process, has received considerable theoretical and empirical attention, leading to a number of hypothesized models (e.g., Cass, 1979; Chapman & Brannock, 1987; Fassinger & Miller, 1996; Minton & McDonald, 1984; Morris, 1997; Rosario, Hunter, Maguen, Gwadz, & Smith, 2001; Troiden, 1989). Although most of these models describe a relatively linear process of identity formation and integration, researchers have recently begun to examine the diverse paths by which LGB individuals become aware of and accept their sexual identity (Floyd & Stein, 2002; Maguen, Floyd, Bakeman, & Armistead, 2002; Rust, 1993; Savin-Williams & Diamond, 2000). Of particular interest has been the coming-out process of lesbian and bisexual women (Chapman & Brannock, 1987; Morris, 1997; Schneider, 2001) because some have suggested that the process is especially variable among women (Diamond, 2000; Kitzinger & Wilkinson, 1995; Peplau & Garnets, 2000; Rust, 1993; Schneider, 2001). Although variability has been observed in the coming-out process of lesbian and bisexual women (Diamond, 2000, 2005; Kitzinger & Wilkinson, 1995; Rust, 1993), the reasons for this diversity have been little explored.

One potential explanation for individual variability in the coming-out process of women may be differences in butch/femme identification. Although the phenomenon of butch and femme has long-standing cultural and historical significance in the lesbian and bisexual female community (Kennedy & Davis, 1989, 1992; MacCowan, 1992; Nestle, 1992) that may have biological foundations (Brown, Finn, Cooke, & Breedlove, 2002; Singh, Vidaurri, Zambarano, & Dabbs, 1999), potential butch/femme differences in the coming-out process remain under examined. The current study investigated a number of hypothesized differences between butch and femme young women on aspects of sexual identity formation and integration in order to further our understanding of the role that butch/femme identification may play in the progression through the coming-out process.

Definition of Butch and Femme

Despite the widespread discussion of butch and femme in various popular and academic works on lesbian and bisexual women, little consensus exists on the definition of what constitutes butch/femme identity. Butch women have been characterized as traditionally masculine in their personality and femme women as traditionally feminine. However, research has found that butch and femme women do not differ on measures of masculine or feminine personality (Singh et al., 1999). Similarly, Loulan (1990) found in her large survey of lesbians that although butch women were more likely to rate themselves as “masculine,” “aggressive,” and “dominant” and femme women were more likely to rate themselves as “feminine,” the women did not differ on more than 51 other adjectives used to assess masculinity/femininity. Qualitative research with smaller samples has suggested that butch and femme lesbian and bisexual women have reported that stereotypical notions of butches as masculine, assertive, and dominant, and femmes as feminine and passive, while sometimes accurate, are often inaccurate (Levitt, Gerrish, & Hiestand, 2003; Levitt & Hiestand, 2004).

Although butch/femme identity may not be associated with masculine or feminine personality characteristics, butch/femme is often characterized by differences in physical appearance. There are acknowledged differences in dress and hairstyle, such that some butch women are more likely to be mistaken for men and femme women are more likely to be perceived as heterosexual women (Loulan, 1990). Some have suggested that the more masculine dress and hairstyle of butches may serve functional purposes to visually identify them as lesbians (Cogan, 1999; Krakauer & Rose, 2002), attract femme partners (Levitt & Hiestand, 2004), and allow them to feel most comfortable and true to themselves (Levitt & Hiestand, 2004). Similarly, femmes report that their feminine appearance makes them feel most comfortable and serves to attract and please a butch partner (Levitt et al., 2003).

Consistent with physical appearance, butch/femme identity may be associated with differences in sexual partner preference, such that butches prefer femme partners and vice-versa. Some research has noted butch/femme differences in partner preference (Levitt & Hiestand, 2005; Pearcey, Docherty, & Dabbs, 1996), but other research has failed to find such differences (Bailey, Kim, Hills, & Linsenmeier, 1997; Smith & Stillman, 2002). Although Loulan (1990) found that 40% of butches are more attracted to femmes and 45% of femmes prefer butches, 24% of butches were more attracted to other butches and 35% of femmes preferred other femmes. Further, Bailey et al. (1997) found that lesbians, regardless of their own butchness or femmeness, preferred more femme-looking than butch-looking partners, but no preference was indicated for femme-acting or butch-acting partners.

Perhaps related to partner preference, butch/femme identity has also been described as a sexual role in which butch women are the active sexual partner. Indeed, more butches rated themselves sexually as “tops” and more femmes rated themselves “bottoms” (Singh et al., 1999). Although butches are generally more likely than femmes to be the sexual initiator (57% vs. 29%; Loulan, 1990), this does not imply that butches are always the active partner. Indeed, Loulan (1990) found that 60% of femmes and 59% of butches reported that they liked being the receptive partner. Differences have also been noted in active/passive sexual dynamics, such that femme women enjoy being actively flirtatious, whereas butch women take pleasure in being the object of this flirtation (Levitt & Hiestand, 2005). Further, femme women report sexual pleasure in being the object of desire, whereas butch women take pleasure in pleasing their partner (Levitt & Hiestand, 2005).

Prevalence of Butch and Femme

Clearly, the phenomenon of butch and femme is multifaceted and complex such that no universal rules define butch/femme identity for all women. Further, the estimates of the prevalence of butch/femme identity are problematic because nearly all research is this area has relied on small convenience samples. However and regardless of the specific reasons for or correlates of butch/femme identity, there is little dispute that butches and femmes compose a significant portion of the lesbian and bisexual female community. Although the prevalence of self-identification as butch or femme has fluctuated over time, perhaps because of the stigmatization of butch/femme by the feminist movement of the 1970s and 1980s (Crawley, 2001; Kennedy & Davis, 1989; MacCowan, 1992; Nestle, 1992), the persistence of butch and femme in the United States, despite this stigma, and the existence of butch and femme cross-culturally (Whitam, Daskalos, Sobolewski, & Padilla, 1998), both implicate biological factors potentially underlying butch/femme identity (Brown et al., 2002; Pearcey et al., 1996; Singh et al., 1999) and underscore the need to understand this characteristic of lesbian and bisexual women.

Although composing a significant portion of the lesbian and bisexual female community, the prevalence of butch/femme identification varies with the assessment method. When women are offered butch, femme, or androgynous response choices, 33% to 60% have identified as butch or femme (Levitt & Horne, 2002; Loulan; 1990; Weber, 1996). A higher prevalence of butch/femme identification has been found when women were forced to choose only between butch and femme (e.g., 85% identified as butch or femme and 15% refused; Brown et al., 2002). Others have conceptualized all lesbian and bisexual women as having aspects of butch and femme and, as such, have asked women to rate themselves on the degree to which they are butch or femme (e.g., Loulan, 1990; Singh et al., 1999). When using such a measure, only 26% of women rated themselves as equally butch and femme (i.e., truly androgynous) and 73% rated themselves as at least somewhat butch or femme (Loulan, 1990).

The Coming-Out Process of Butch and Femme Women

Despite a growing literature on the sexual identity formation and integration of lesbian and bisexual women (e.g., Diamond, 2000; Kitzinger & Wilkinson, 1995; Rust, 1993), this literature has not provided information regarding potential butch/femme differences in the coming-out process. Further, the growing literature on butch/femme differences in biology and sexual relationships (Bailey et al., 1997; Brown et al., 2002; Levitt & Hiestand, 2005; Pearcey et al., 1996; Singh et al., 1999) has not examined potential differences in the coming-out process. Nevertheless, some largely qualitative research has provided evidence to suggest that the coming-out process may differ between butches and femmes (Hiestand & Levitt, 2005; Levitt et al., 2003; Levitt & Hiestand, 2004; Levitt & Horne, 2002).

One traditional marker of the coming-out process concerns the age at initiating various milestones of sexual identity development (e.g., Floyd & Stein, 2002; Maguen et al., 2002; Rosario et al., 1996; Rust, 1993; Savin-Williams & Diamond, 2000). In the sole examination of psychosexual milestones among butches and femmes (Levitt & Horne, 2002), butches were found to become aware of their sexual attraction to women at an earlier age (M = 14.6 years, SD = 11.0) than femmes (M = 21.9 years, SD = 10.6), although the difference was only marginally significant. In addition, butches have been noted to develop more gender atypical behaviors (e.g., tomboys) at a very early age (Hiestand & Levitt, 2005), whereas femmes have reported feeling that their childhoods were easier (i.e., less stigmatized, less isolated) than those of butches because they were unaware of their sexual orientation until later (Levitt et al., 2003). The current study extended this literature by examining butch/femme differences on a large set of psychosexual milestones of sexual identity development, including the timing of same-sex attractions, behavior, and sexual identity. We hypothesized that the greater visibility of young butches (i.e., gender atypicality) will lead them to question their sexual orientation, initiate same-sex behavior, and identify as lesbian or bisexual earlier than young femmes.

Another aspect of the coming-out process is the unfolding of sexual identities, sexual attractions, and sexual behaviors that are congruent with one another (Rosario, Schrimshaw, Hunter, & Braun, 2006). Thus, a young woman who adopts a lesbian identity would be expected to be more same-sex oriented in her attractions and behaviors than a woman who is consistently bisexual or who may be questioning her sexual identity. This process necessarily takes time, often years. Over time, those women who are questioning their sexual identity and some of those who identified as bisexual will become lesbian identified and more same-sex oriented in their attractions and behaviors, whereas other women who are exploring their sexual identity will find that a bisexual identity is best attuned to their sexual attractions and behavioral preferences (Rosario et al., 2006). Regardless, if butches do initiate the coming-out process earlier than femmes, as hypothesized above, young butch women as a group would be expected to be more same-sex oriented in their identity, attractions, and behaviors than young femme women (some of whom may still be questioning and exploring their sexual identity) because butches have had a longer time to develop their sexual identity. These hypothesized differences in sexual self-identification, sexual orientation, and behavior were examined in this report.

The coming-out process also involves the integration of a lesbian or bisexual identity into the woman’s life. This includes accepting that sexual identity, becoming involved in LGB-related social activities, resolving internalized homophobia, becoming more comfortable with others knowing about their sexual identity, and disclosing that identity to others (Morris, 1997; Rosario et al., 2001). However, potential butch/femme differences in sexual identity integration remain little examined. There is some indication that butches and femmes may differ on self-disclosure of their sexual identity as lesbian or bisexual. Butches have described that, because of their butchness, some parents and teachers suspected they would be lesbians and, thus, were not surprised when butches self-disclosed this sexual identity (Hiestand & Levitt, 2005). By comparison, femmes reported having a harder time disclosing because heterosexuals often incorrectly assumed that they were heterosexual and other lesbians believed they were too feminine to identify as lesbian (Levitt et al., 2003). Therefore, we hypothesized that young butches will have disclosed their sexual identity to more individuals, will be more comfortable with others knowing their sexual identity, and will have experienced less internalized homophobia than young femmes.

Finally, if butch/femme differences in sexual identity exist (i.e., femmes are more likely to be bisexually identified, oriented, etc.), this raises the question of whether observed butch/femme differences in the coming-out process are due to the butch/femme phenomenon or to the greater prevalence of bisexuality among femmes. Given the potential confound between butch/femme and sexual identity, previously noted elsewhere (Carr, 2005), as well as the well-documented differences between lesbian and bisexual women in the coming-out process (Diamond, 2000; Rosario et al., 1996; Rust, 1993), the current study examined the interaction between butch/femme identity and sexual identity in order to eliminate this possible confound. We compared lesbian femmes, bisexual femmes, lesbian butches, and bisexual butches in order to investigate whether variability in the coming-out process was a function of differences in butch/femme identity and/or in sexual identity as lesbian or bisexual.

In summary, the current study examined potential butch/femme differences in various aspects of sexual identity formation and integration of young lesbian and bisexual women because the phenomenon of butch and femme may be a factor that contributes to the variability in the coming-out process of lesbian and bisexual women.

METHOD

Participants

As part of a larger study, male and female youth between the ages of 14 and 21 years were recruited from organizations serving LGB youth in New York City, including three LGB-focused community-based organizations (CBOs) and two LGB student organizations from public colleges. Most youth (85%) were recruited from the three CBOs. Of the 164 youth interviewed at baseline, 49% were female. The current report focuses on these lesbian and bisexual female youth. Of the 81 young women interviewed, 1 was excluded because she was older than the study’s age criterion and 4 were excluded because they identified as heterosexual and had never had sex with someone of the same sex.

The final sample consisted of 76 female youth with a mean age of 18.4 years (SD = 1.6). The youth were of Latina (38%), Black (36%), White (20%), Asian (3%), and other (4%) ethnic backgrounds. Although 16% of the young women were recruited from college organizations, 32% were in college. Thirty-eight percent of the female youth reported that they had a parent who received welfare, food stamps, or Medicaid (defined here as “low” socioeconomic status). Youth self-identified as lesbian (67%), bisexual (32%), or other (1%).

Procedure

Voluntary and signed informed consent was provided by all youth. For youth under age 18 years, parental consent was waived by the Commissioner of Mental Health for New York State. An adult at each CBO served in loco parentis to safeguard the rights of each minor-aged research participant. The university’s Institutional Review Board and the recruitment sites approved this study.

Youth completed a 2- to 3-hour interviewer-administered questionnaire at baseline and subsequent assessments 6 and 12 months later. Baseline interviews were conducted from October 1993 through June 1994, with follow-up interviews conducted through August 1995. Interviewers were college-educated individuals of the same sex as the youth. Interviewers were trained and received weekly supervision. Youth received $30 at each interview.

Only 4 (5%) female youth were lost to both follow-up assessments. The retention rates for the female youth were 87% (66/76) for the 6-month assessment and 88% (67/76) for the 12-month assessment.

Measures

Butch and Femme Identity

The classification of the young women as butch and femme was based on three separate assessments: (1) the youth’s self-identification as butch/femme, (2) the youth’s report of how other lesbians would identify her, and (3) the interviewer’s rating of the youth. At the baseline, 6-month, and 12-month assessments, youth were asked to respond to three items regarding whether or not “you think of yourself or see yourself as” butch, femme, and/or androgynous (in that order). Youth responded yes or no to each item. The test-retest reliability of this measure over a two-week period was assessed for 30 young women at the baseline assessment (97% consistent for butch, κ = .87; and 87% consistent for femme, κ = .73). Because butch, femme, and androgynous identifications were assessed separately, young women could identify with just one identity or with multiple identities. This was done for two reasons. First, because of the stigma attached to butch/femme identity, we believed that allowing for multiple identities might be less stigmatizing than forcing a single identification. Further, because there is acknowledged variability within butch and femme identified women (such as women who identify as “soft butch,” who are on the androgynous side of butch), assessing multiple identities allowed us to capture the full spectrum of butch and femme women, not just those who would identify as “pure” butch or femme. Consistent with this rationale, for every assessment period, youth who identified only as butch or as butch and androgynous were defined as butch. Similarly, youth who identified only as femme or as femme and androgynous were defined as femme. Youth who identified as butch and femme, or as butch, femme, and androgynous, or only as androgynous, were defined as androgynous. Finally, youth who did not identify as butch, femme, or androgynous (i.e., they responded negatively to all three items) were defined as having none of these identities. Therefore, a single variable was computed at each assessment period composed of four possible responses as butch (3), femme (2), androgynous (1), and none of the above (0).

Because we anticipated that some young women might be reluctant to self-identify as butch or femme (due to stigma or discomfort with that identity), we also asked youth how other lesbians would perceive them, so as to allow them to indirectly classify themselves. Youth were asked, “Would you say that on average other lesbians see you or react to you as” butch, femme, and/or androgynous (in that order). Again, the youth provided a yes or no response to each of these three items at each assessment. Test-retest reliability of the young women’s reports of how other lesbians viewed them was assessed (80% consistent for butch, κ = .44; and 87% consistent for femme, κ = .73) over two weeks. Following the same classification system outlined above, we computed a single variable at every assessment period consisting of four possible responses as butch (3), femme (2), androgynous (1), and none of the above (0).

To avoid youth’ potential reluctance with self-identifying as butch/femme, we also obtained interviewer ratings at the 12-month assessment. Interviewers responded to a single item asking whether, in their opinion, the youth would be likely to “be perceived by average casual observers, such as people on the street” as butch (3), femme (2), or androgynous (1). No test-retest reliability data were available for this measure because the interviewer ratings were made only at the 12-month assessment.

As expected, the three measures of butch/femme were correlated. There was 58%-59% agreement (Contingency Coefficient [CC] = .62-.67) between the self-identification as butch/femme and others’ perceptions of her as butch/femme at each of the three assessments. Similarly, the interviewer ratings at the 12-month assessment were also similar to the self-identification (54% agreement, CC = .63) and the others’ perceptions of her (54% agreement, CC = .56) at the same time period. Therefore, these three methods of assessment were combined into a single composite measure of butch/femme identity for each time period. Specifically, the youth’s self-identification and the youth’s perceptions of how other lesbians would perceive her were combined, within each assessment period, by using the maximum value as the indicator of the youth’s butch/femme identity [as coded above, in which butch = 3, femme = 2, androgynous = 1, and none of the above = 0]. For the 12-month assessment, we also included the interviewer’s ratings in computing the indicator of the youth’s butch/femme identity. Three composite butch/femme variables were computed, one for each assessment period, representing the youth’s identity as butch (3), femme (2), androgynous (1), or none of the above (0). These three composite variables were also correlated over time, such that there was a 66%-77% agreement (CC = .61-.72) among time periods. Therefore, we combined the three composite variables from each time period into a single measure of butch/femme identity by using the maximum value of the youth’s identity across the three assessment periods as the indicator of the youth’s butch, femme, or androgynous identity.

Sexual Identity and Sexual Identity Formation

Assessment of sexual identity, sexual developmental milestones, sexual orientation, and sexual behaviors were assessed using the Sexual Risk Behavior Assessment-Youth (SERBAS-Y) for LGB youth (Meyer-Bahlburg, Ehrhardt, Exner, & Gruen, 1994). The SERBAS-Y has demonstrated strong test-retest reliability among lesbian and bisexual female youth (Schrimshaw, Rosario, Meyer-Bahlburg, & Scharf-Matlick, 2006). Each component of the SERBAS-Y used in this report is discussed in detail below.

Psychosexual Developmental Milestones

The SERBAS-Y assessed the ages (in years) when the young women experienced various milestones in the development of sexual attraction, identity, and behavior. Youth were asked the ages when they were first attracted to, fantasized about, and were aroused by erotica focusing on women. The ages reported for these three milestones were correlated (.68 < r < .75) and a factor analysis of these three items generated a single factor (Cronbach’s α = .88), suggesting that they occurred at the approximately the same age. Therefore, mean age of these three milestones was computed to obtain a single age when these three events occurred, resulting in a variable indicating the age of awareness of same-sex sexual orientation. Comparable items assessing opposite-sex attractions, fantasies, and erotic arousal were similarly combined (Cronbach’s α = .89). Young women also were asked about the age when they first thought they “might be” lesbian/bisexual and when they first thought they “really were” lesbian/bisexual. Finally, the young women were asked about the age when they first engaged in any one of several specific sexual activities (e.g., manual, vaginal-digital, oral, vaginal-penile, and analingus) with women and with men. The minimum age reported across the various possible behaviors with female or male partners was used as the age when the youth first had any sex with women and the age when they first had any sex with men.

We agree with others (e.g., Diamond & Savin-Williams, 2000) who have critiqued sexual milestone research suggesting that the ages, per se, do not tell us much about the coming-out process, as individuals may first experience these milestones at any time in life. However, because the coming-out process necessarily takes time to work through and integrate, we have argued that more important than age when milestones occur is the how recently the milestones occurred (i.e., the length of time between first experiencing these milestones and the present; Rosario et al., 2006). Indeed, two young women who first experienced attractions towards women at age 15 may actually be very different, if one is currently 16 (one year later) and the other is 21 (six years later). Thus, for all of the developmental milestones, we computed the number of years since the youth first experienced the various milestones by subtracting the age at each milestone from the youth’s age at the baseline assessment.

Lifetime and Recent Sexual Behaviors

At the baseline assessment, items from the SERBAS-Y assessed the lifetime prevalence of various sexual behaviors, including the number of male and female sexual partners and the number of male and female sexual encounters. After defining the various sexual behaviors to be assessed (i.e., manual, vaginal-digital, oral, vaginal-penile, and analingus), the youth were asked to “count up” all of their female sexual partners with whom they had “any kind of sex with in their whole lifetime.” This was followed by a question regarding the number of times they had sex with these female partners. Similar questions were then asked regarding sexual behaviors with male partners. In addition, the gender of recent sexual partners was assessed. At the baseline interview, youth were asked if they had female or male sexual partners in the past three months. In the two subsequent assessments, youth were asked about the gender of their partners in the past 6 months (i.e., since the last interview). For our analyses of recent sexual activity, we focused on whether youth reported any sexual activity with women or men.

Current Sexual Orientation

The SERBAS-Y was used to assess current cognitive sexual orientation at every assessment period by means of three Kinsey-style items. Youth were asked the extent to which their recent sexual attractions, thoughts, and fantasies were focused on women or men (1) when in the presence of other individuals, (2) while masturbating, dreaming, or day dreaming, and (3) when viewing erotic material in films, magazines, or books. A 7-point Likert response scale was used ranging from “always guys/men” (0) to “always girls/women” (6), with a midpoint (3) indicating equally focused on both sexes. Youth were given the option of indicating they had none of the assessed experiences. Current cognitive sexual orientation was computed as the mean of the three items (Cronbach’s α = .91-.92 across the three assessments).

Sexual Identity

A single item from the SERBAS-Y assessed sexual identity at every assessment period by asking youth, “When you think about sex, do you think of yourself as lesbian, bisexual, or straight?” Youth rejecting such identities were coded as “other” and asked to elaborate.

Sexual Identity Integration

Involvement in Gay-Related Activities

The prevalence of lifetime involvement in gay/lesbian-related social activities was assessed at baseline using a 28-item scale developed for this study (Rosario et al., 2001). At subsequent assessments, youth were asked about their involvement in the past 6 months (i.e., since their last assessment). A factor analysis of the baseline data of the full sample of youth identified 11 items (e.g., going to a gay bookstore, gay coffee house, gay pride march, gay clubs or bars) that loaded on a single factor. The number of these items endorsed by the youth was used as the indicator of involvement in gay-related activities (Cronbach’s α = .64-.77 across the three assessments).

Positive Attitudes Toward Homosexuality/Bisexuality

A 33-item scale adapted from the Nungesser Homosexual Attitudes Inventory (Nungesser, 1983) was administered at all three assessments using a 4-point response scale ranging from “disagree strongly” (1) to “agree strongly” (4). Factor analysis of the baseline data identified two factors in the full sample. The first factor contained 11 items (e.g., “My [homosexuality/ bisexuality] does not make me unhappy”) that assessed attitudes toward homosexuality/bisexuality. The mean of these items was computed at each assessment with a high score indicating more positive attitudes toward homosexuality/bisexuality (Cronbach’s α = .83-.85 across the three assessments). Because the youth’ attitudes were negatively skewed at all assessments (e.g., M = 3.59 of a maximum possible value of 4.0, SD = 0.48 at baseline), the data were transformed using the exponential e to stretch the positive end of the distribution.

Comfort with Others Knowing about Your Homosexuality/Bisexuality

A modified version of the Nungesser Homosexual Attitudes Inventory (Nungesser, 1983) was administered at all three assessments. As noted above, a factor analysis of the baseline data identified two factors. The second factor contained 12 items (e.g., “If my straight friends knew of my [homosexuality/bisexuality], I would feel uncomfortable”) that assessed comfort with others knowing the youth’s sexuality. The mean of these items was computed for each time period, with a high score indicating more comfort with homosexuality/bisexuality (Cronbach’s α = .89-.91 across the three assessments).

Self-Disclosure of Sexual Identity to Others

Youth were asked at baseline to enumerate “all the people in your life who are important or were important to you and whom you told that you are (lesbian/bisexual)” (Rosario et al., 2001). At subsequent assessments, youth were asked about the number of individuals to whom they had disclosed during the past six months (i.e., since the last assessment). Because disclosure cannot be undone and because it is carried into the future, the indicator of disclosure is cumulative over time. Therefore, we summed the disclosure data at baseline with new disclosures at the 6- and 12-months assessments, such that the total number of individuals disclosed to up to that point was used as our indicator of self-disclosure. Because the follow-up data were positively skewed, the scores for the 6- and 12-month assessments were logarithmically transformed after the raw data were summed.

Certainty About, Comfort With, and Self-Acceptance of Sexuality

At the 6-month and 12-month assessments, items were added to assess the commitment of the youth to their lesbian identity or to that part of their bisexual identity that was centered on the same sex (Rosario et al., 2006). We asked youth who self-identified as lesbian, “How certain are you about being lesbian at this point?” and asked the bisexual youth, “How certain are you about being bisexual at this point?” For comfort with sexuality, we asked the lesbian youth, “How comfortable are you with your lesbianism?” and asked the bisexual youth, “How comfortable are you with your lesbian side?” For self-acceptance of sexuality, we asked the lesbian youth, “How accepting of your lesbianism are you?” and asked the bisexual youth, “How accepting are you of your lesbian side?” We coded the prevalence of being very certain/ comfortable/ accepting (1) as compared to being less than very certain/comfortable/accepting (0).

Other Measures

Social Desirability

The Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1964) was administered at baseline to assess the tendency to provide socially desirable responses. Two items that were inappropriate for youth were removed, resulting in a 31-item scale, which was administered using the original true-false response format. As with other research on LGB youth (Safren & Heimberg, 1999), a factor analysis of the full sample identified 12 items that loaded on a single factor. The number of these items endorsed was computed as the indicator of social desirability (Cronbach’s α = .74).

Data Analysis

In order to describe the prevalence of butch/femme identity of the sample at each time period, we computed the percentage of youth who self-identified as butch/femme, who reported other lesbians would perceive her as butch or femme, and the interviewer ratings of the youth as butch or femme (at the 12-month assessment only). Bivariate comparisons of butches and femmes on aspects of sexual identity formation and integration were made using ANOVA for continuous variables (e.g., time since first experiencing a developmental milestone, comfort with others knowing their sexual identity) and chi-square for categorical variables (e.g., percentage of those who reported same-sex partners, percentage who reported being certain of their sexual identity). Post-hoc comparisons were made using Fisher’s Protected t-test for continuous variables and chi-square for categorical variables. Measures of effect size (i.e., the amount of variance explained) were computed using eta-square (η²) for continuous variables and Goodman and Kruskal’s (1979) tau (τ) for categorical variables. In addition, multivariate comparisons of butches and femmes (in which controls were imposed for significant covariates) were made using multiple linear regression for continuous outcomes and multiple logistic regression for categorical outcomes. For the multivariate analyses, effect size measures are reported, specifically beta (β) for linear regression and the odds ratio (OR) for logistic regression. Identical statistical procedures were used in the comparison of the intersection between butch/femme identity and sexual identity (e.g., lesbian butches, lesbian femmes, bisexual femmes). All statistical tests reported were significant at p < .05, unless otherwise indicated.

RESULTS

Classification and Prevalence of Butch and Femme

Table I provides, for each time period, the percentage of butch, femme, or androgynous youth, based on self-reported butch/femme identity, how youth thought other lesbians would perceive them, and how they were classified by the interviewer. Although highly correlated (see Measures section) prevalence of butch/femme differed by these three assessment methods. Youth were more likely to report being perceived by other lesbians as butch or femme (72%-75% across the time periods) than they were to self-identify as butch or femme (40%-52%). This difference was especially noticeable for butch identity. Specifically, youth were more likely to report that other lesbians would perceive them as butch than they were to self-identify as butch at all three time periods (e.g., 26% vs. 12%). Similarly, fewer youth identified as butch (9%) than were so perceived by the interviewer (28%).

Table I.

Prevalence of Butch and Femme by Three Types of Measurement

	Self-Identification	How Other Lesbians Perceive You	Interviewer Rating1
	N (%)	N (%)	N (%)
Baseline
Butch	9 (12%)	20 (26%)	NA
Femme	30 (40%)	37 (49%)	NA
Androgynous	16 (21%)	11 (15%)	NA
None of the Above	21 (28%)	8 (11%)	NA
6-month assessment
Butch	10 (15%)	18 (27%)	NA
Femme	22 (33%)	30 (46%)	NA
Androgynous	18 (27%)	13 (20%)	NA
None of the Above	16 (24%)	5 (8%)	NA
12-month assessment
Butch	6 (9%)	17 (26%)	14 (28%)
Femme	21 (31%)	30 (46%)	14 (28%)
Androgynous	23 (34%)	14 (21%)	22 (44%)
None of the Above	17 (25%)	5 (8%)	NA

Note. NA = not assessed.

¹Interviewer ratings were made at the 12-month assessment only. Failure of the interviewers to make ratings for every youth resulted in missing data.

Because the youth’ self-reported identity tended to minimize their butch/femme identity, the data from the three assessment methods were combined in order to classify each youth (see Methods for details). This categorization procedure identified 33 (43%) butches, 39 (51%) femmes, 4 (5%) androgynous, and 0 youth who did not identify with any of these labels. Because there were too few youth with an androgynous identity to conduct statistical analyses, all subsequent analyses compared just butch (1) and femme (0) young women (n = 72).

Comparisons of Butches and Femmes

Potential Covariates

Butches and femmes were similar on sociodemographic characteristics and social desirability. No significant differences (p < .10) were found between them on socioeconomic status, race/ethnicity, recruitment site, or social desirability. However, butches were older than femmes (r = .25, p < .05) by almost one year (M = 18.8 and 18.0 years old, respectively).

Sexual Identity Formation

Bivariate comparisons between butches and femmes on aspects of sexual identity formation are presented in Table II. However, because butches and femmes differed significantly on age, we examined these differences controlling for age in multivariate analyses using linear regression for continuous outcomes and logistic regression for dichotomous outcomes. The bivariate and multivariate findings were highly similar, and thus only the multivariate findings are discussed here. In the multivariate analyses, butches and femmes did not differ significantly on the number of years since first experiencing any of the various psychosexual milestones. However, significant differences were consistently found in sexual behaviors. Butches reported more lifetime female partners (β = .27) and encounters (β = .33) than femmes, whereas femmes reported more lifetime male partners (β = −.30) and encounters (β = −.32) than butches. Similar findings were noted in the lifetime prevalence of sex, with butches more likely than femmes to report sex with women (OR = 6.5, p = .09) and less likely to report sex with men (OR = .32, p = .09). There were also differences on more recent sexual behaviors. At each of the three assessments, butches were more likely than femmes to be sexually involved with women (OR = 2.7-9.6 across the three assessments), but butches were less likely to be sexually involved with men (OR = .10-.20). In addition, the cognitive sexual orientation of butches was more strongly focused on women than was that of femmes at all three assessments (β = .38-.45).

Table II.

Differences in Psychosexual Milestones, Sexual Behaviors, and Sexual Orientation by Butch/Femme Identity

	Butch (n = 33)	Femme (n = 39)	Statistical Test	Effect Size
Psychosexual Milest ones	M (SD)	M (SD)	F	η 2
Years since first:1
Same-sex sexual attractions, fantasies, and arousal	7.6 (3.0)	6.7 (3.3)	1.32	.02
Other-sex sexual attractions, fantasies, and arousal	6.6 (1.8)	6.2 (3.1)	0.45	.01
Thought might be lesbian or bisexual	5.9 (3.7)	5.9 (4.3)	0.00	.00
Other-sex sexual activity	6.5 (3.5)	4.7 (3.5)	3.61†	.06
Same-sex sexual activity	4.4 (3.3)	4.9 (4.8)	0.22	.00
Thought really was lesbian or bisexual	3.7 (2.2)	3.0 (3.1)	1.38	.02
Sexual Behaviors	M (SD)	M (SD)	F	η 2
Lifetime # female partners1,2	12.8 (34.9)	3.8 (4.8)	9.44**	.12
Lifetime # male partners1,2	4.7 (9.4)	6.8 (7.7)	4.54*	.06
Lifetime # female sexual encounters1,2	709.3 (1232.3)	210.9 (665.8)	12.19***	.15
Lifetime # male sexual encounters1,2	43.6 (87.0)	153.5 (235.8)	5.76*	.08
	%	%	χ 2	τ
Lifetime Prevalence of Sex with Women1	97%	80%	5.00*	.07
Lifetime Prevalence of Sex with Men1	70%	87%	3.31†	.05
Recent Sex with Women
Time 1	79%	54%	4.91*	.07
Time 2	86%	65%	3.54†	.06
Time 3	93%	69%	5.90*	.09
Recent Sex with Men
Time 1	6%	39%	10.40***	.15
Time 2	14%	44%	6.43**	.10
Time 3	14%	46%	7.52**	.12
Cognitive Sexual Orientation 3	M (SD)	M (SD)	F	η 2
Time 1	5.5 (0.9)	4.3 (1.5)	17.21***	.20
Time 2	5.4 (0.8)	4.2 (1.8)	11.93***	.17
Time 3	5.3 (1.0)	4.2 (1.7)	10.50**	.15

¹Variable assessed at Time 1 only.

²Means and SDs reported were based on untransformed data, but statistical tests were conducted with logarithmically transformed data.

³The scale can range from 0 to 6.

^†p < .10.

^*p < .05.

^**p < .01.

^***p < .001.

Sexual Identity Integration

Bivariate comparisons between butches and femmes on aspects of sexual identity integration are presented in Table III. Controlling for age in multivariate analyses revealed an identical pattern of findings. As compared with femmes, butches consistently indicated that they were more comfortable with others knowing about their homosexuality (β = .32-.50), were more likely to self-identify as lesbian than bisexual (OR = 10.0-29.7), and were more comfortable with their sexual identity (OR = 4.9-9.2, p < .07). In addition, butches were significantly involved in more gay-related activities at 12 months (β = .41), endorsed more positive attitudes toward homosexuality at 6 months (β = .28), and were more likely to be accepting of their sexual identity at 12 months (OR = 6.6, p < .09) than femmes, although these findings were not significant at all three assessment periods.

Table III.

Differences in Aspects of Sexual Identity Integration by Butch/Femme Identity

	Butch (n = 33)	Femme (n = 39)	Statistical Test	Effect Size
Sexual Identity Integration Process	M (SD)	M (SD)	F	η 2
Involvement in gay/lesbian social activities1
Time 1	7.1 (2.4)	6.1 (2.7)	2.70	.04
Time 2	6.3 (2.5)	5.3 (2.7)	1.95	.03
Time 3	6.7 (3.2)	4.3 (2.6)	11.21***	.15
Positive attitudes toward homosexuality/bisexuality2
Time 1	46.0 (9.7)	42.9 (11.9)	1.44	.02
Time 2	46.6 (9.3)	39.2 (12.9)	6.52**	.10
Time 3	44.3 (9.8)	43.0 (13.4)	0.20	.00
Comfort with others knowing of her homosexuality/bisexuality3
Time 1	3.1 (0.7)	2.7 (0.6)	8.09**	.11
Time 2	3.2 (0.5)	2.7 (0.6)	12.63***	.17
Time 3	3.4 (0.6)	2.8 (0.6)	14.23***	.19
Self-disclosure to others4
Time 1	7.3 (4.9)	6.1 (4.2)	1.37	.02
Time 25	2.4 (1.2)	2.3 (1.0)	0.03	.00
Time 35	2.7 (1.3)	2.6 (1.0)	0.19	.00
	%	%	χ 2	τ
Self-identify as lesbian
Time 1	91%	46%	15.54***	.22
Time 2	100%	48%	19.66***	.35
Time 3	96%	53%	14.32***	.24
Certainty of sexual identity
Time 2	82%	63%	2.54	.05
Time 3	93%	77%	2.71	.05
Comfort with sexual identity
Time 2	93%	73%	3.80*	.07
Time 3	96%	77%	4.54*	.08
Acceptance of sexual identity
Time 2	93%	85%	0.93	.02
Time 3	96%	77%	4.54*	.08

¹The scale may range from 0 to 11.

²The scale may range from 2.72 to 54.60, given the exponential transformation.

³The scale may range from 1 to 4.

⁴The scale was open-ended, and therefore can range from 0 to infinity.

⁵The disclosure measure at Time 2 and Time 3 were logarithmically transformed; therefore they are not comparable with the untransformed data from Time 1.

^*p < .05.

^**p < .01.

^***p < .001.

Prevalence of Butches and Femmes by Self-Identification as Lesbian or Bisexual

Although the findings reported so far support the hypothesized differences between butches and femmes, these observed differences may conflate butch/femme identity and sexual identity as lesbian or bisexual. As observed earlier (Table III), dramatic differences existed between butch and femme youth on sexual identity with nearly all butches (91%-100%) identifying as lesbian but nearly half of femmes identifying as bisexual (46%-53%). This suggests that butch/femme differences may be confounded with sexual identity. Thus, to examine whether the observed butch/femme differences were attributed to differences in sexual identity, we treated butch and femme identity as distinct from sexual self-identification as lesbian or bisexual when examining the four (2 × 2) mutually exclusive groups that butch/femme identity by sexual identity might generate. There were 29 lesbian butches, 18 lesbian femmes, 21 bisexual femmes, and 3 bisexual butches. Given the small number of bisexual butches, less than what would be expected by chance, we excluded them from further analyses. Therefore, all subsequent analyses compared lesbian butches, lesbian femmes, and bisexual femmes.

Comparison of Butches and Femmes by Self-Identification as Lesbian or Bisexual

Potential Covariates

Lesbian butches, lesbian femmes, and bisexual femmes were compared on sociodemographic characteristics and social desirability. They did not differ significantly on SES, race/ethnicity, recruitment site, or social desirability. However, the groups differed on age, F (2, 65) = 3.58, p < .05, with pairwise comparisons indicating that lesbian butches were significantly older than both lesbian and bisexual femmes (M = 19.0, 18.0, and 18.1 years, respectively).

Sexual Identity Formation

Bivariate comparisons among lesbian butches, lesbian femmes, and bisexual femmes on aspects of sexual identity formation are presented in Table IV. Given the significant differences among these three groups, multivariate analyses also were conducted in which age was controlled. In the multivariate analyses, only one significant difference was found in the psychosexual milestones. Specifically, lesbian femmes first experienced same-sex attractions a longer time ago than did bisexual femmes (β = .33).

Table IV.

Differences in Milestones, Sexual Behaviors, and Sexual Orientation by Butch/Femme Identity and Sexual Identity

	Lesbian Butch (n = 29)	Lesbian Femme (n = 18)	Bisexual Femme (n = 21)	Statistical Test	Effect Size
Psychosexual Milestones	M (SD)	M (SD)	M (SD)	F	η 2
Years since first1
Same-sex sexual attractions,  fantasies, and arousal	7.7 (3.1)a	8.0 (2.9)a	5.6 (3.2)b	3.79*	.10
Other-sex sexual attractions,  fantasies, and arousal	6.8 (1.8)	6.6 (3.7)	5.9 (2.7)	0.61	.03
Thought might be lesbian or bisexual	6.2 (3.8)	6.1 (4.5)	5.8 (4.2)	0.08	.00
Other-sex sexual activity	6.5 (3.5)	5.6 (4.1)	4.0 (3.0)	2.62	.09
Same-sex sexual activity	4.5 (3.2)	3.9 (4.4)	6.1 (5.3)	1.11	.04
Thought really was lesbian or bisexual	3.9 (2.3)	3.5 (3.4)	2.5 (2.7)	1.54	.05
Sexual Behaviors	M (SD)	M (SD)	M (SD)	F	η 2
Lifetime # female partners1,2	14.1 (37.1)a	5.8 (5.8)a	2.1 (2.9)b	10.32***	.24
Lifetime # male partners1,2	3.4 (5.0)a	4.9 (5.6)a	8.4 (9.0)b	4.00*	.11
Lifetime # female sexual encounters1,2	800.3 (1290.2)a	436.9 (942.8)	17.1 (8.4)b	19.69***	.38
Lifetime # male sexual encounters1,2	44.3 (91.7)a	66.6 (141.5)a	228.0 (275.6)b	7.50***	.19
	%	%	%	χ 2	τ
Lifetime Prevalence of Sex with Women1	97%a	94%a	67%b	10.73**	.16
Lifetime Prevalence of Sex with Men1	69%a	78%	95%b	5.18†	.08
Recent Sex with Girls/Women
Time 1	79%a	83%a	29%b	17.45***	.26
Time 2	88%a	81%a	50%b	8.49*	.14
Time 3	92%a	88%a	53%b	11.31**	.19
Recent Sex with Boys/Men
Time 1	7%a	0%a	71%b	35.21***	.52
Time 2	16%a	25%a	61%b	10.28**	.17
Time 3	12%a	19%a	68%b	18.12***	.30
Cognitive Sexual Orientation 3	M (SD)	M (SD)	M (SD)	F	η 2
Time 1	5.5 (0.9)a	5.5 (0.6)a	3.3 (1.3)b	39.23***	.55
Time 2	5.5 (0.7)a	5.4 (0.6)a	3.0 (1.7)b	29.58***	.51
Time 3	5.3 (1.0)a	5.3 (0.8)a	3.3 (1.7)b	17.41***	.38

Note. Means or prevalences with differing superscripts differed significantly at p < .05.

¹Variable assessed at Time 1 only.

²Means and SDs were based on untransformed data, but statistical tests were conducted with logarithmically transformed data.

³The scale can range from 0 to 6.

^†p < .10.

^*p < .05.

^**p < .01.

^***p < .001.

As compared with the psychosexual milestones, a consistent pattern of differences was found with regard to sexual behavior and cognitive sexual orientation, in which lesbian butches and lesbian femmes did not differ significantly, but both differed from bisexual femmes. Specifically, lesbian butches (β = .47) and lesbian femmes (β = .39) reported more lifetime female partners than bisexual femmes, and lesbian butches (but not lesbian femmes) reported fewer lifetime male partners than bisexual femmes (β = −.46). Similarly, lesbian butches (β = .62) and lesbian femmes (β = .55) reported more sexual encounters with women than did bisexual femmes; both groups also reported fewer sexual encounters with males than bisexual femmes (β = −.55 and −.38 respectively). Likewise, lesbian butches (OR = 10.5) and lesbian femmes (OR = 10.7) were more likely than bisexual femmes to have ever had sex with a women, whereas lesbian butches (OR = 0.10) were less likely than bisexual femmes to have ever had sex with a man. Consistent with these lifetime sexual behaviors, both lesbian butches (OR = 5.8-19.0 over the three assessments) and lesbian femmes (OR = 4.6-12.5, p < .06) were more likely to report recent sexual involvement with women than were bisexual femmes. Conversely, lesbian butches (OR = 0.03-0.11) and lesbian femmes (OR = 0.09-0.21) were less likely than bisexual femmes to report recent sexual involvement with men. In addition, the cognitive sexual orientation of lesbian butches (β = .66-.81) and lesbian femmes (β = .58-.70) was much more centered on women than was the cognitive sexual orientation of bisexual femmes.

Sexual Identity Integration

Bivariate comparisons of the lesbian butches, lesbian femmes, and bisexual femmes on aspects of sexual identity integration are presented in Table V. Further, multivariate analyses controlling for age indicated that lesbian butches generally differed significantly from bisexual femmes, whereas lesbian femmes were often intermediate between these two groups. Indeed, lesbian femmes sometimes differed significantly from bisexual femmes and at other times did not differ significantly from either lesbian butches or bisexual femmes. Specifically, both lesbian butches (β = .30-.49 across the three assessments) and lesbian femmes (β = .32 at baseline only) reported more involvement in gay/lesbian social activities than bisexual femmes, but did not differ from each other. Similarly, both lesbian butches (β = .56 at Time 2 only) and lesbian femmes (β = .37-.52 at Times 2 and 3) reported more positive attitudes toward homosexuality than bisexual femmes, but did not significantly differ from each other. Lesbian butches (β = .32 at Time 1 only) and lesbian femmes (β = .28-.29 at Time 1 and 2) also reported greater disclosure of their sexuality to others than did bisexual femmes. In contrast, lesbian butches reported more comfort with others knowing about her sexuality than both lesbian femmes (β = .27-.41, p < .08) and bisexual femmes (β = .38-.59) at all three assessments. However, on some aspects of sexual identity integration, differences between butch lesbians and bisexual femmes were noted, but lesbian femmes did not differ significantly from either group. Specifically, lesbian butches were more certain about their sexual identity (OR = 8.5-10.0) and more comfortable with their sexual identity (OR = 11.3-21.1) than bisexual femmes, but lesbian femmes did not differ significantly from either group.

Table V.

Differences in Aspects of Sexual Identity Integration by Butch/Femme Identity and Sexual Identity

	Lesbian Butch (n = 29)	Lesbian Femme (n = 18)	Bisexual Femme (n = 21)	Statistical Test	Effect Size
Sexual Identity Integration Process	M (SD)	M (SD)	M (SD)	F	η 2
Involvement in gay/lesbian social activities
Time 1	7.3 (2.2)a	7.1 (2.2)a	5.2 (2.8)b	5.10**	.14
Time 2	6.1 (2.6)	6.0 (3.0)	4.7 (2.3)	1.76	.06
Time 3	6.6 (3.3)a	4.9 (2.2)	3.7 (2.8)b	5.40**	.16
Positive attitudes toward homosexuality/bisexuality
Time 1	45.5 (10.3)	46.4 (8.6)	39.9 (13.6)	2.13	.06
Time 2	47.2 (9.5)a	46.6 (8.6)a	32.2 (12.5)b	13.01***	.31
Time 3	44.1 (10.2)	48.4 (9.3)a	38.4 (14.7)b	3.31*	.10
Comfort with others knowing of her homosexuality/bisexuality
Time 1	3.1 (0.7)a	2.8 (0.6)	2.6 (0.6)b	4.19*	.12
Time 2	3.2 (0.5)a	2.8 (0.6)b	2.6 (0.5)b	7.03**	.20
Time 3	3.3 (0.6)a	2.9 (0.6)b	2.7 (0.5)b	6.62**	.19
Self-disclosure to others
Time 1	7.4 (4.9)a	7.7 (4.9)a	4.7 (3.0)b	2.92†	.08
Time 25	2.5 (1.2)	2.7 (0.9)	2.0 (1.1)	1.98	.06
Time 35	2.8 (1.3)	2.9 (0.8)	2.3 (1.1)	1.95	.06
	%	%	%	χ 2	τ
Certainty of sexual identity
Time 2	88%a	75%	46%b	7.38*	.14
Time 3	96%a	87%	67%b	6.26*	.12
Comfort with sexual identity
Time 2	92%a	88%	50%b	9.10*	.18
Time 3	96%a	100%	53%b	16.82***	.31
Acceptance of sexual identity
Time 2	92%	94%	70%	4.01	.08
Time 3	100%a	87%	67%b	9.09*	.17

Note. Means or prevalences with differing superscripts differed significantly at p < .05.

¹The scale may range from 0 to 11.

²The scale may range from 2.72 to 54.60, given the exponential transformation.

³The scale may range from 1 to 4.

⁴The scale was open-ended, and therefore can range from 0 to infinity.

⁵The disclosure measure at Time 2 and Time 3 were logarithmically transformed; therefore, they are not comparable with the untransformed data from Time 1.

^†p < .10.

^*p < .05.

^**p < .01.

^***p < .001

DISCUSSION

Despite considerable interest in the variability observed in the coming-out process of lesbian and bisexual women, little research has examined butch/femme identification as a potential explanation for the diversity of women’s developmental experiences. Building on earlier and primarily qualitative research suggesting butch/femme differences (e.g., Hiestand & Levitt, 2005; Levitt et al., 2003; Levitt & Horne, 2002), the current report examined potential differences in the sexual identity formation and integration of butch and femme young women.

No satisfactory definition of butch and femme exists because potential defining aspects, such as masculine and feminine personality characteristics, appearance, partner preferences, and sexual behavior roles are not as strongly related to butch/femme identity as one would expect. Although this may suggest that butch/femme identity is a multidimensional rather than a unitary phenomenon, its multidimensionality poses assessment difficulties for classifying butch and femme individuals. Although our three measures of butch/femme (i.e. self-identification, perceptions of how other lesbians might view them, and interviewer assessment of how individuals might perceive the youth) were correlated, suggesting that they did assess a similar underlying construct, there were also discrepancies among the measures. These discrepancies may be attributed to the reluctance of women to self-identify as butch or femme. Our finding that some young women did not self-identify as butch or femme, although they believed that other lesbians would so identify them and the interviewer also concurred that many of the young women would be so identified by other individuals, documents this reluctance to self-identify. Regardless, reluctance to self-identify as butch or femme requires that researchers use alternative strategies to assess butch/femme identity.

The reluctance to self-identify as butch or femme and, thus the exclusive reliance on self-identification as butch or femme, presents significant problems for research on butch/femme identity. Although a woman may not identify as butch or femme, her experiences may be that of a butch or a femme, if other individuals perceive and relate to her as if she were one or the other. Furthermore, reluctance to self-identify implies that the prevalence of butch and femme is likely to be underestimated. The measurement imprecision of exclusively relying on self-identification may result in nonsignificant empirical relations between butch/femme self-identification and hypothesized outcomes (e.g., the coming-out process). This could generate an inconsistent literature of contradictory significant and nonsignificant findings attributed to measurement error rather than theoretical misspecification.

In an attempt to overcome the reluctance to self-identify as butch or femme, we used three sources of information to classify youth. Although our measurement solution resulted in a greater prevalence of young women classified as butch/femme, it is unclear if the women who self-identified as butch/femme were similar to the young women who were indirectly classified as butch/femme, given insufficient power to compare the two groups. By combining these two groups (in order to overcome the stigma of self-identification), we may have reduced our ability to detect butch/femme differences, if it is only self-identified butches and femmes who differ. However, our significant and hypothesized findings of butch/femme differences and differences between lesbian butches and bisexual femmes suggest that our composite assessment of butch/femme was valid and may be useful in future research.

We examined hypotheses linking butch/femme identity with sexual identity development, including both sexual identity formation and integration. Initially, support was found for the hypothesis that the sexual identity development of butches would be further along than that of femmes. However, we found that butches were nearly exclusively lesbian-identified and femmes were equally likely to identify as lesbian or bisexual. This suggested that butch/femme identity and lesbian/bisexual identity might be conflated. Moreover, the conflation could generate invalid impressions about butches and femmes. The identification of confounding between butch/femme identity and sexual identity is an important consideration for future research on butch/femme identity as it suggests that this research must also consider sexual identity. Although some have avoided this issue by focusing only on butch/femme differences among lesbian-identified women (excluding bisexual women), this may not be desirable when examining the coming-out process. To the extent that bisexually-identified women may be in the early process of forming a lesbian identity, the exclusion of bisexuals may be problematic because it would eliminate women from various phases of the coming-out process and, in so doing, potentially mislead our understanding of lesbians’ sexual identity development. Rather, our finding suggests that future research should seek to simultaneously examine butch/femme identity and sexual identity. This would require forethought and planning to ensure that there was a sufficient sample to examine all four possible groups of butch/femme by sexual identity.

When butch/femme identity and sexual identity were disentangled, we found the young women were classified primarily into three groups: lesbian butches, lesbian femmes, and bisexual femmes. Because our examination found few bisexual butches, additional research is needed to ascertain whether bisexual butches compose a valid category and how they may differ from lesbian butches, lesbian femmes, and bisexual femmes. We expect that bisexual butches do not constitute a prevalent group because, although there may be exceptions, bisexual women may be more likely to attempt to maintain a more femme presentation so as to maximize their attractiveness to both male and female partners. Further, bisexual women, some of whom may be less certain of their identity, less comfortable with their identity, and less comfortable with others knowing about their identity (e.g., Rosario et al., 2006), may maintain a more femme self-presentation so as to avoid others questioning whether they are lesbian or bisexual. Likewise, young butch women who identify as bisexual may be more likely to be questioned about their sexuality than femme-identified bisexuals, perhaps leading the former to question their sexual identity earlier than femme-identified women.

Comparisons of our three groups found relatively few differences between lesbian butches and lesbian femmes on sexual identity formation, but both groups differed from bisexual femmes. For example, lesbian butches and femmes reported more female partners and a cognitive sexual orientation that was more focused on women than did bisexual femmes. Such findings are expected, if sexual identity formation is differentiated by and, therefore, best understood by lesbian and bisexual identity. This failure to identify differences between lesbian femmes and lesbian butches suggests that sexual identity formation is not particularly influenced by butch/femme. This is unsurprising, if identification as a butch or femme lesbian occurs after identification as a lesbian in general (Hiestand & Levitt, 2005). If butch/femme identity development follows sexual identity formation, then butch/femme differences in sexual identity formation would not be expected.

With respect to sexual identity integration, the interaction of butch/femme identity by sexual identity was important. Lesbian butches once again differed from bisexual femmes on integration. Lesbian femmes generally did not differ significantly from lesbian butches, except that lesbian femmes were consistently more uncomfortable with others knowing about their homosexuality than were butches. Lesbian femmes sometimes differed significantly from bisexual femmes on sexual identity integration. However, more often than not, lesbian femmes did not differ significantly from bisexual femmes on sexual identity integration (61%, 11/18 multivariate comparisons). Therefore, lesbian butches, lesbian femmes, and bisexual femmes compose three groups, with lesbian butches and bisexual femmes being most distinct and lesbian femmes overlapping somewhat with one or the other group depending on the issue under examination.

As we considered sexual identity integration, in which community and social factors would be expected to exert strong influences, we were uncertain of the pattern of findings that might emerge. Lesbian femmes might cope with the stigma attached to lesbianism by avoiding mention of and drawing attention away from their lesbianism. Thus, they would “pass” or allow others to incorrectly infer that they were heterosexual. If lesbian femmes so coped, their sexual identity integration would be less than that of butches and perhaps similar to that of bisexual femmes, if these women also engaged in passing. On the other hand, as lesbian femmes embraced their sexuality, they would become more similar to lesbian butches on sexual identity integration. Our findings, which found that lesbian femmes were sometimes like lesbian butches and sometimes like bisexual femmes supported both of these perspectives. This may indicate that our young lesbian femmes were in the process of integrating their sexual identity, but remained conflicted on some integration aspects. Longitudinal studies beyond the one-year follow-up period of the current study would be needed to track the sexual identity integration of lesbian femmes relative to that of lesbian butches in order to determine which set of hypotheses is supported. If lesbian femmes were in the process of sexual identity integration and/or of resolving sexual identity conflicts, they eventually would become similar to lesbian butches on sexual identity integration. If lesbian femmes and lesbian butches were to compose distinct groups, the differences observed here might persist over time.

The data also have implications for understanding the coming-out process of bisexual femmes. The findings that bisexual femmes differed consistently from lesbian butches and sometimes from lesbian femmes may be partly attributed to their bisexual identity rather than to any observed quantitative differences in the coming-out process. Indeed, we found that bisexual femmes were more likely to report sex with men, to have more male partners, and to report greater sexual attractions for men. Although some of these factors may suggest sexual exploration for some individuals, all of them may also indicate a consistent and confident bisexual identity for other individuals. Further, although bisexual femmes were found to be less involved in gay and lesbian social activities, there is some question about whether bisexually-identified women want to be involved in “gay and lesbian” social activities. Thus, this difference on involvement may be due to the women’s bisexuality, and not reflect the degree to which they have progressed through the coming-out process.

Nevertheless, and as already implicated, other aspects of sexual identity integration assessed here—attitudes toward one’s own homosexuality/bisexuality, comfort with others knowing about one’s homosexuality/bisexuality, disclosure of sexual identity to others, and certainty, comfort, and acceptance of one’s sexuality—should be equally relevant to both self-identified lesbian and bisexual women. Our finding that lesbian butches differed from bisexual femmes on these aspects of sexual identity integration, with lesbian femmes sometimes differing from one group and sometimes from the other group, suggest three things. First, they support the conclusion of an interaction between butch/femme by lesbian/bisexual differences in the coming-out process. Second, they indicate that the young bisexual femmes in this study, and to a lesser degree, the young lesbian femmes, continue to work through the issues involved in sexual identity integration. Third, the findings suggest that the butch/femme phenomenon may help us understand some of the variability that has been found in the sexual identity development of lesbian and bisexual women.

One final comment about the findings is in order. Although we found substantial numbers of differences on sexual behavior, sexual orientation, and sexual identity integration, we found few differences in sexual identity formation. Specifically, we found only one significant difference in psychosexual milestones out of 12 possible comparisons. This general pattern of nonsignificant differences indicated that time since first experiencing various developmental milestones of identity formation did not differ by butch/femme identity or by lesbian/bisexual identity. Rather, the observed differences were found in more recent behaviors and experiences, suggesting that the differences emerged subsequent to sexual identity formation. This is perhaps because the timing of sexual identity formation may be determined by biopsychological processes that are relatively constant across individuals, whereas sexual identity integration may be more socially and culturally influenced (Rosario, Schrimshaw, & Hunter, 2004).

The present research has limitations requiring mention. First, our sample size was modest because our analyses focused on only the young female youth in our study (49% of the full sample). Although we had adequate statistical power for the relatively simple bivariate and multivariate analyses conducted here, we encourage future research with larger samples. Second, our sample may not be representative of all young lesbian and bisexual women because it was recruited from gay-focused organizations or programs. In essence, our young women may be further along in the coming-out process than youth recruited from non-gay focused venues. However, our sample of young women is not likely to be as far along on the coming-out process as samples of butch and femme adult women, who are often recruited from gay-related organizations or gay festivals. Also, because the sample was recruited from New York City and because it is significantly more racially and ethnically diverse than those of past studies, it is unclear how our findings may generalize to older, more suburban or rural, and less diverse samples. Finally, the study was limited to a one-year follow-up period. However, the coming-out process is a long-term developmental process, which may explain why there was relatively little change in aspects of the coming-out process observed in this report. Therefore, future research should employ longer follow-up periods so that changes in the coming-out process can be examined. This report provides the justification for longer-term follow-up studies.

Despite these limitations, this report provides critical empirical data on potential butch/femme differences in the coming-out process among lesbian and bisexual young women that may serve to further understand the diversity of sexual identity development found among lesbian and bisexual women.