Careless responding in online student learning surveys: profiles and impact on data quality

Investigations into the prevalence of careless responding (CR) in online student learning survey research are scarce, and little to no information is present about the impact on data quality. This study illustrates how an unobtrusive post hoc method of CR detection can be applied using four indices, which can be calculated in most student learning surveys. By applying latent profile analyses in an exemplary sample of 1,111 first-year university students, different responder profiles are identified, showing the presence of careful and careless responder profiles. The study further examines the impact of CR on data quality, raising concerns about reliability and validity. The study calls for more attention to the presence of CR when analysing data in online student learning survey research. CR detection is also valuable for practice, specifically when online survey-based feedback is provided to learners with more questionable responding behaviour.