Confidentiality Assurances and Response: A Quantitative Review of the Experimental Literature

This article tests the general hypothesis that a stronger assurance of confidentiality improves survey response by means of a meta-analysis of the experimental literature. No support is found for the general hypothesis, but the subsidiary hypothesis, that confidentiality assurances improve response when the data asked about are sensitive, is supported. Under those circumstances, the effect of confidentiality assurances is small but statistically significant and is robust in the presence of a variety of control variables.