Forced‐response in online surveys: Bias from reactance and an increase in sex‐specific dropout

Due to computer technology, a forced‐response can be easily achieved in online questionnaires and is frequently used to gather complete datasets. An Internet‐based quasi‐experiment was conducted on the student server at the University of Vienna to study the influence of forced‐response on dropout, demographic reports, and the content of the results. Forced‐response was shown to substantially increase dropout. In addition, forced‐response interacted with reported sex in eliminating a naturally occurring sex difference in dropout that was observed for the questionnaire whenever responses did not need to be enforced. Also reported sex turned out to have a mediating effect on time of dropout: Men dropped out earlier than did women. Further analyses revealed a reactance effect, as predicted by reactance theory. It is concluded that data from online questionnaires with forced‐response designs are in danger of being hampered by dropout and reactance.