Methodological Issues in the Effects of Attrition: Simple Solutions for Social Scientists

Abstract

Participant attrition from longitudinal research studies is a concern for social scientists because loss of certain subgroups of participants may result in subsequent data collection phases becoming increasingly biased. This article examines whether attrition has been given due consideration in selected reports of longitudinal research undertaken by social scientists. Results of a review of the literature found that less than one-quarter of the studies described how data were examined for patterns of attrition. On the other hand, those articles that described the treatment of attrition used solutions that were varied and often simple and effective. Recommendations for statistical and nonstatistical ways of dealing with sample attrition used by social scientists are provided.