Do required fields in online surveys in the social sciences impair reliability?

ABSTRACT

The present study is the first to examine empirically whether required fields in online surveys impair reliability and response pattern, as participants forced to respond to all items may provide arbitrary answers. Two hundred and thirteen participants completed a survey consisting of six questionnaires testing personal and social issues and perceptions. They were randomly assigned to one of two versions of the survey: optional-fields (N = 104) or required-fields (N = 109). Comparison of the Cronbach’s alpha of the two versions revealed identical reliability values for all questionnaires, save for somatization, where a minor difference was found. Confirmatory factor analysis showed no difference in the factor structure of the two versions, and no differences were found by Bayesian t-test and Levene’s test for equality of variances. The findings suggest that required fields do not impair reliability or change the response pattern, and therefore can be used in online surveys to prevent missing data.