An assessment of the effect of data collection setting on the prevalence of health risk behaviors among adolescents.

Abstract

To examine the effect of data collection setting on the prevalence of priority health risk behaviors among adolescents. Analyses were conducted using data from two national probability surveys of adolescents, the 1993 national school-based Youth Risk Behavior Survey (YRBS) and the 1992 household-based National Health Interview Survey (NHIS/YRBS). Forty-two items were worded identically on both surveys. Thirty-nine of the 42 identically worded items (93%) showed that the YRBS produced estimates indicating higher risk than the NHIS. Twenty-four of these comparisons yielded statistically significant differences. The prevalence estimates affected most were those for behaviors that are either illegal or socially stigmatized. School-based surveys produce higher prevalence estimates for adolescent health risk behaviors than do household-based surveys. Each has advantages and disadvantages, and both can play a role in assessing these behaviors.