Relevance & Research Question:
Drop outs - people leaving a survey before completing it – are a very common problem in online studies. Whereas, some design-related measures to reduce drop out-rates have been implemented (for an overview see Fan & Yan, 2010), it is to date unclear what kind of internal factors of participants (e.g., personality, motivation) affect drop out. In the present study we analyzed whether the Big Five personality factors and participant’s choice of reimbursement has an effect of respondents motivation to continue filling out a web survey.
Methods & Data:
N = 3,013 individuals (71.6% women, mean age M = 23.4 years, SD = 3.9 years) took part in an online-study. After choosing the preferred kind of reimbursement (deciding to be paid by participating in a lottery or through a personality feedback), participants filled out a brief measure of the Big Five. To provoke drop out, several large personality questionnaires followed divided into eight blocks of 40-60 items each. Completing all 460 questions of the whole study took participants on average 63.0 minutes (SD = 21.4 minutes). Using discrete-time survival analysis, the effect of the five personality dispositions and choice of reimbursement on drop out was investigated.
Results:
Only 736 participants (24.4%) completed the whole online-survey. We found that drop out was influenced by individual’s openness, their agreeableness and the their choice of reimbursement: People who were more open to experiences, more agreeable, as well as those who chose to get personality feedback instead of participating in a lottery were less likely to quit participation. Importantly, the effect of all three variables was mediated by satisfaction with the questionnaire in the previous questions block. That is, all three variables lead to a higher satisfaction which in turn lowered the probability to quit the survey.
Added Value:
While we have good knowledge which design-related measures helps us to reduce drop out, little is known about personality aspects that could lead to survey non-response. Our study shows which Big Five personality aspects moderate drop out. Thus, relevant personality aspects are identified and should thus be additionally considered in future studies.