A Latent Profile Analysis of Bisexual Identity: Evidence of Within-Group Diversity.

Bisexual people experience unique psychosocial vulnerabilities and their mental health needs and social identity remain underserved and understudied, respectively. We report results from a latent profile analysis where we identified a preliminary typology of bisexual identity subgroups and its association with demographic and mental health variables. Bisexual+ adults (N = 292) residing in the U.S. were recruited from Amazon Mechanical Turk and administered a demographic survey, indicators of bisexual identity, and measures of internalizing symptoms and self-esteem. Joint consideration of statistical and substantive criteria in the modeling process yielded a well-differentiated and qualitatively distinctive three-profile solution comprised of Affirmative (e.g., having a positive orientation towards one's bisexuality), Vigilant (e.g., being significantly concerned about others' reactions to one's bisexuality), and Ambivalent (e.g., endorsing mixed but generally negative attitudes and beliefs about one's bisexuality) profiles of bisexual identity. Auxiliary analyses revealed conceptually and statistically significant associations among profile membership, demographic covariates, and mental health outcomes. Some key findings included that compared to the Affirmative profile, men and people of color were overrepresented in the Ambivalent profile, whereas men were overrepresented in the Vigilant profile. Bisexuals with a Vigilant profile displayed the poorest mental health constellation. Our findings highlight the categorically heterogeneous nature of bisexual identity, support the relevance of social identity to mental health among bisexuals, and represent the first attempt to model bisexual identity using mixture techniques. Future studies should consider larger and more demographically diverse samples, address replicability and generalizability, examine additional auxiliary variables, and investigate longitudinal developments in profiles.