“Coming out” among gay Latino and gay White men: implications of verbal disclosure for well-being

Abstract

We examined how verbal disclosure affects subjective well-being among gay Latino and White men. In Study 1, increased gay identification predicted increased verbal disclosure for gay White men but not for gay Latino men, who reported verbal disclosure levels in line with a tacit coming out strategy (expressing one’s sexual orientation to others more non-verbally) regardless of their gay identification. Moreover, low verbal disclosure hindered subjective well-being only for gay White men. In Study 2, we show that increased intrinsic self-expression and higher relational self-construal explained the positive relationship between verbal disclosure and well-being among gay White men. This mediational model did not hold for gay Latino men. We discuss verbal disclosure in relation to mainstream understandings of coming out.

Keywords:

• Gay men
• verbal disclosure
• coming out
• tacit subjectivity
• sexual orientation
• well-being

Notes

1. Tacit subjectivity takes its name from the Spanish linguistic concept of el sujeto tacito (the tacit subject) where the subject of a sentence is known through the conjugation of the accompanying verb but is not explicitly stated.

2. We took several steps to maximize the number of recruited MTurk workers who self-identified as gay Latino or gay White men. First, we created two HITs on MTurk (i.e., announcements to participant in a research study). One HIT specifically requested participants to identify as a gay Latino man, and one HIT specifically requested participants to identify as a gay White man. This request was made at the end of the title of the HIT and was also embedded in the description of the HIT several times. We also included an “honesty prompt” at the end of our study. This honesty prompt explained to MTurk workers that it is difficult to receive responses from men of various ethnicities and sexual orientations and that their specific HIT was designed to recruit gay Latino men (or gay White men). Participants were then asked to indicate their gender, ethnicity, and sexual orientation a second time, with the understanding that they would still receive payment even if they did not identify with the identities that the HIT attempted to recruit. Six participants indicated they did not identify with the identities their respective link attempted to recruit. Of these, four gay White participants completed the link meant for gay Latinos but were retained. Thus, only two participants did not fit our inclusion criteria to participant and were, thus, filtered out of analyses.

3. We mistakenly left out one item from the OI. The item left out was “leaders of my religious community.” However, Study 2 includes the full list of items and replicates the effects that involve the verbal disclosure variable.

4. We also ran additional analyses where we examined the participant ethnicity × gay identification interaction for each of the subscales of the Outness Inventory measure (outness to family, outness to the world, and outness to religion). We found comparable patterns in each case. When using outness to family (verbal disclosure to family members), outness to the world (verbal disclosure to friends/acquaintances), and outness to religion (verbal disclosure to members of one’s religious group) as separate criterion variables, increased gay identification predicted increased verbal disclosure for gay White men, bs ≥ 1.10, ts ≥ 3.70, ps < .01, but not for gay Latino men, bs ≤ .43, ts ≤ 1.65, ps > .10.

5. We took several steps to be certain that the MTurk workers for Study 2 were unique and had not completed Study 1. First, MTurk workers reported the last six characters of their unique MTurk worker identification number at the end of Studies 1 and 2 (this also served as their “code” to receive payment via MTurk). Second, after data collection was completed for Study 2, we compared the worker identification numbers reported at the end of Studies 1 and 2 to determine whether an MTurk worker completed both studies. We also downloaded the identification numbers associated with all MTurk workers who completed both studies from MTurk and used this list to aid the comparison task. All worker identification numbers were immediately deleted after the comparison task. Six participants were dropped because they completed both studies.

6. We did not assess gay identification in Study 2 because the focus was on how verbal disclosure related to the two proposed mediators and well-being, particularly for gay Latino men.

7. We also tested whether perceived intrinsic self-expression led to higher relational self-construal and accounted for the effect of verbal disclosure on subjective well-being for gay White men (a sequential mediation model). We used Model 6 in PROCESS (5000 bootstrapped resamples) where we regressed subjective well-being on verbal disclosure and entered intrinsic self-expression and relational self-construal as sequential mediators, only using data from gay White men. Higher verbal disclosure predicted higher levels of intrinsic self-expression, which in turn predicted higher relational self-construal. Higher relational self-construal led to more subjective well-being. Consequently, the sequential indirect effect of verbal disclosure on subjective well-being through intrinsic self-expression and relational self-construal for gay White men was significant, indirect effect = .03, SE = .02, 95% CI = [.00, .10]. Interestingly, when switching the order of the mediators, so that relational self-construal preceded intrinsic self-expression, the sequential indirect effect was also significant, indirect effect = .03, SE = .02, 95% CI = [.00, .09].