Use of a Respondent-Generated Personal Code for Matching Anonymous Adolescent Surveys in Longitudinal Studies.

PURPOSE: Research on sensitive and private topics relies heavily on self-reported responses. Social desirability bias may reduce the accuracy and reliability of self-reported responses. Anonymous surveys appear to improve the likelihood of honest responses. A challenge with prospective research is maintaining anonymity while linking individual surveys over time.

METHODS: We have tested a secret code method in which participants create their own code based on eight questions that are not expected to change.

RESULTS: In an ongoing middle school trial, 95.7% of follow-up surveys are matched to a baseline survey after changing up to two-code variables. The percentage matched improves by allowing up to four changes (99.7%).

CONCLUSIONS: The use of a secret code as an anonymous identifier for linking baseline and follow-up surveys is feasible for use with adolescents. While developed for violence prevention research, this method may be useful with other sensitive health behavior research.