Int Conf AIDS. 1996 Jul 7-12;11(1):180 (abstract no. Mo.D.1733). Unique
Objective: To establish the reliability of sexual behavior self-reports
in a sample of 75 male couples with mixed HIV-status. Methods: First we
interviewed couple's members separately and queried them on their sexual
behavior with each other during the previous year. Second, we
interviewed them together and discussed the differences, if any, in
their selfreports. After one week we re-interviewed 27 of the couples.
Test-retest reliability analyses were conducted on the 27 individuals,
and inter-partner reliability was explored for the whole sample.
Results: In terms of test-retest reliability of frequency of performance
of specific sexual practices (measured on a five point Likert scale),
the weighted Kappas varied from 0.7 (C.I. 0.5-0.9) to .09 (C.I.
0.8-1.0), thus exhibiting consistency over time. Test-retest of number
of sexual occasions was also quite reliable (.9 Intra-class
correlation). Interpartner correlations for number of sexual occasions
was good 7 Inter-class correlation). However, interpartner agreement
concerning shared sexual practices was poor, with Kappa values ranging
from 0 (C.I. -0.1-0.2) to 0.4 (C.I. 0.2-0.6). Length of relationship was
associated with number and magnitude of discrepancies, the shorter the
relationship the larger the differences. Most inter-partner
discrepancies were due to confusion concerning the long recall period
(one year prior to assessment), and lack of recall of infrequent
behaviors. Conclusions: These data point out that although individuals
may be consistent in their self-reports of sexual behavior, their
information may not agree with that of the partners with whom the sexual
acts took place. Shortening recall periods and analyzing infrequent
behaviors separately may be a way of minimizing unreliability.
*Communication *Homosexuality, Male *Sex Behavior *Sexual Partners