Improving estimates of prevalence rates of sensitive behaviors: The randomized lists technique and consideration of self‐reported honesty

Publisher Website
Google Scholar
Abstract
Although self‐report data are generally accepted by behavioral scientists, there continues to be concern about the validity and reliability of self‐reports of sensitive behaviors, such as sexual behaviors, drug use, and criminal behavior. Our aim for the current study was to present an example of the use of the Randomized Lists Technique (RLT, a variation of the Randomized Response Technique) in conjunction with self‐reported honesty in assessing and improving the validity of prevalence estimates of sexual behaviors with a sample of high school students. Both techniques allow respondents to provide information to the researcher about a sensitive behavior without having to report the behavior directly. Study participants included 1,886 tenth grade students in the Dade County (Greater Miami) Public Schools. The RLT‐estimated prevalence rate was 80% higher for sex with a same‐gender partner than the result for the direct question. Inconsistencies in reporting sex with a same‐gender partner or having had intercourse were generally much higher for respondents who said they were “not very honest” than those who said they were “completely honest.” Our findings indicate that the combination of the RLT and weighting by reported honesty may yield the most accurate estimates of these behaviors. The likelihood of underreporting was reduced by the RLT, and overreporting appears to have been reduced by the use of honesty weights.