Sensitive topics: Are there modal differences?

As organizations such as the US Navy move toward web-based survey administration, it is important to determine whether responses to sensitive questions are equivalent regardless of the administration method. It is possible that having to enter a military website and provide a username to access the survey might increase anonymity and confidentiality concerns, thereby increasing the likelihood of response distortion. Although previous research has shown that self-administered questionnaires (both paper- and computer-based) generally result in equivalent responses, to date little research has investigated the effects of asking sensitive information in a web-based survey format. In order to test for potential response bias, a survey was created containing questions about family planning attitudes and birth control usage, impression management and self-deception scales, and privacy and confidentiality concerns. Paper surveys were sent to about 3300 randomly selected enlisted Navy personnel while another 3300 received an invitational letter to a survey website. In general, the results show few differences between web-based and paper-based survey responses, demonstrating that sensitive data can be collected via web-based surveys without significantly influencing the honesty of survey responses. In addition, the perceived levels of privacy and confidentiality did not result in a significant difference between the responses provided.