A Reexamination of the Bogus Pipeline

Abstract

A variant of a technique developed to enhance the validity of self-reports, the generalized bogus pipeline (GBP), was investigated in three experiments reported in this paper. Empirical support for both the construct validity and the efficacy of the GBP was generally supportive for that manipulation. An important serendipitous finding, however, was that the GBP may be unnecessary when respondent anonymity is assured. Empirical support for the construct validity and efficacy of anonymity assuring procedures was much stronger. Meta-analyses across all three experiments suggests that the effect size of anonymity in obtaining more valid self-reports of sensitive behavior was larger than the effect size of the GBP and that the GBP did little to improve the veracity of self-reports beyond the effect of anonymity. Implications of the findings are discussed in terms of the relative merits of the bogus pipeline paradigm and anonymity procedures with suggestions to increase the validity of self-reports utilizing anonymity assuring procedures.