Who Are They and Why Don't They Finish?: Answering Drop-outs

Abstract Web-based online surveys are now a well-recognized method,of gathering data. Their use allows tracking of a particular kind of non-respondent ‐ t he answering dropout, who begins answering questions but drops out before completing the surve y. Converting this group into completers may be a relatively easy way to increase response rates . In order to do so, the factors relating to their dropout must be understood. This study examines responses to surveys to determine which survey and respondent factors may be related to dro pping out of surveys. Survey responses from ten surveys (eight for instructors and two for stud ents) administered in the Cisco Networking Academy Program were analyzed. For each survey, number of questions, number of pages, median time to complete, number of each type of questions, number of topics covered, and translation were coded. For each respondent survey completion status, country, computer infrastructure, and education level were coded. Vis ual analysis of graphical patterns suggests rules of thumb to help reduce rates of answering dr opouts.