Methodological challenges in research on sexual risk behavior: II. Accuracy of self-reports

Abstract

Assessing sexual behavior with self-report is essential to research on a variety of health topics, including pregnancy and infertility, sexually transmitted infections, and sexual health and functioning. Recent methodological research has provided new insights regarding the accuracy of self-reports of sexual behavior. We review these studies, paying particular attention to a promising new development: the use of computer-assisted assessments. The collection of sexual risk behavior data with computers has increased dramatically in recent years, but little is known about the accuracy of such assessments. We summarize the evidence, discuss methodological issues that arise in studies evaluating the accuracy of self-reports, and offer recommendations for future research.

Test–Retest Reliability of Self-Reported Sexual Behavior, Sexual Orientation, and Psychosexual Milestones Among Gay, Lesbian, and Bisexual Youths

“…The rationale for the use of kappa and ICC over Pearson or Spearman correlations have been argued elsewhere (e.g., Schroder et al, 2003). Briefly, whereas interclass correlations (e.g., Pearson, Spearman) are appropriate for examining the relation between two independent variables, these correlations are inappropriate when the two variables share variance (e.g., two assessments of the same variable).…”

“…However, researchers have questioned the ability of individuals to provide reliable self-reports and whether current measurement strategies reliably assess sexual information (Catania, Gibson, Chitwood, & Coates, 1990;Schroder, Carey, Vanable, 2003;Weinhardt, Forsyth, Carey, Jaworski, & Durant, 1998). Failure to reliably assess self-reported sexual information would have profound consequences for research on sexuality and sexual health.…”

“…Indeed, a recent review of the research conducted since 1990 identified only 15 studies that examined test-retest reliability of self-reported sexual behavior (Schroder et al, 2003), with far fewer studies conducted prior to 1990 (Catania et al, 1990). Taken together, research on the reliability of self-reported sexual behavior has been characterized as "a mixed bag" (Catania et al, 1990).…”

“…Taken together, research on the reliability of self-reported sexual behavior has been characterized as "a mixed bag" (Catania et al, 1990). The more recent research continues to have a number of limitations; see Schroder et al (2003) for a recent critique. To adequately evaluate reliability, questions must assess the same behaviors for the same period in time (e.g., the past month) at both the test and retest assessments.…”

“…While social desirability and privacy concerns may reduce the reliability of reported sexual behaviors of all populations (see Catania et al, 1990;Schroder et al, 2003 for review), this may be particularly true among GLB individuals, given the stigma attached to same-sex sexuality. However, few studies have examined the reliability of sexual behavior among GLB individuals.…”