![[ Visit Client Website ]](images/banner.gif)
Methods: In Phase II, we recruited LGBT (n=244) and non-LGBT (n=250) service members for an online quantitative survey, using respondent driven sampling (RDS) within the separate cohorts. RDS is a chain-referral sampling method that assumes hidden populations are better able to access peers (Heckathorn, 1997). It uses a dual-incentive structure where participants receive a primary incentive for their own participation and secondary incentives for recruiting peers.
Thirty-two “seeds” provided by Phase I participants, study investigators, and the advisory panel completed the Phase II instrument and in turn referred additional participants into our study, establishing the first RDS chains. RDS chains were also initiated via strategic advertising in general-population and LGBT-specific online publications and social networking platforms. English-speaking active duty service members, age 18 or older, in one of the four main branches were eligible for participation. Individuals were assigned to the LGBT cohort if they self-identified as LGBT, or if their assigned and asserted genders did not match, and to the non-LGBT cohort if they identified as heterosexual and their asserted gender matched their sex assigned at birth.
Phase II participants completed a battery of demographic and behavioral health measures, described in the accompanying abstracts. Participants were offered $25 for completing the survey during non-duty hours, although n=8 completed the survey during duty hours, foregoing the incentive. After completing the survey, participants established a unique referral code used to recruit up to three other participants. LGBT participants were specifically instructed to refer other LGBT individuals, whereas non-LGBT participants were simply asked to refer any active duty service members. For each referral, the recruiter received an additional $10. Substantial effort was made to eliminate participants and cut off referral chains of suspected duplicates and suspicious or fraudulent data (Robinson-Cimpian, 2014).
Results: Of the 544 valid responses recorded during the Phase II data collection period, 290 (53%) entered the survey via RDS. Fifty-one distinct referral chains emerged from the respondent population. “Cross-over,” the referral of LGBT service members by non-LGBT participants (or vice versa), was just under 10% in both directions. Transgender (F(2,541)=28.17, p<0.001) and sexual minority (LGB) service members (F(2,541)=6.31, p<0.01) were rated significantly less accepted than women or racial/ethnic minorities overall, and post-hoc tests revealed that members of these subgroups rated their group’s acceptance as even lower than non-group members did.
Implications: RDS is a promising strategy for the recruitment of hard-to-reach active duty military personnel, including LGBT service members. Perceptions of acceptance vary by demographic status and may have serious consequences for behavioral health outcomes.