Non-Response Bias with a Web-Based Survey of College Students: Differences from a Classroom Survey About Carrying Concealed Handguns

Internet-based and e-mail surveys represent viable administrative methods for efficiently collecting data. These methods appear to be particularly well-suited for studying college student populations, a group that has gained attention from criminologists interested in testing theories. An important concern with administering surveys with the Internet and via e-mail is that of non-response bias. Despite the appeal of online surveys, nonresponse bias associated with these methods has not been sufficiently investigated. The study described here estimates nonresponse bias associated with a web-administered survey that measured opinions about changing concealed handgun carrying laws on college campuses, items likely to elicit polarizing opinions. Results show important substantive differences between web-administered and in-class versions of the survey. Students who responded to the web survey expressed more extreme opinions and behavioral responses to a proposed policy that would allow concealed handgun carrying on campus. Survey researchers who utilize web-based administrative methods should consider using multiple sources of leverage when soliciting participation and must carefully evaluate sample representativeness.