Mode effects in surveys of gay men: A within-individual comparison of responses by mail and by telephone

In an evaluation of a community-level AIDS prevention program, a random sample of gay bar patrons was sent questionnaires by mail, and a random household sample was interviewed by phone. Over 3 baseline years, 581 members of the bar cohort switched response modes for various reasons, including reassignment by the investigators. Regardless of mode sequence, these same men were roughly twice as likely to report unprotected anal intercourse with a nonprimary partner by mail as by phone, but only half as likely to report volunteering in AIDS efforts. The mode effects found here for gay men are comparable to those previously found among Blacks, in keeping with previous reports of elevated social desirability scores in both groups.