Abstract
Because previous literature has discussed but not scientifically measured the impact that type of separation and identifiability of respondent have on re- sponses to questions during the process of exit interviewing and surveying (EIS), this study sought to examine the impact of these variables on a sample of separating Army personnel. Results showed that although both variables have an impact on responses, the impact is not consistent for either response distor- tion or amount of missing responses. The data are discussed in terms of study limitations and the impact of impression management on EIS responses.