Personality traits, panel tenure, survey topic, and context as predictors of survey nonresponse patterns in high-frequency online longitudinal surveys.

An extensive literature studies the relation between demographic and socio-economic characteristics and attrition in longitudinal studies. In this study, we analyze the independent effects of non-demographic variables-respondent personality traits, panel tenure, and survey topics, using unique datasets from two recently completed high-frequency online longitudinal studies conducted in the U.S. We used latent class analysis to group respondents into various classes based on similarities in their nonresponse patterns across all survey waves, which revealed substantial variations in patterns of nonresponse. Our results indicate that respondent personality traits were strong predictors of nonresponse patterns. Specifically, conscientiousness is positively associated with a lower likelihood of wave nonresponses. In contrast, more open, extroverted, neurotic, and agreeable respondents are more likely to exhibit higher wave nonresponses, but with effect sizes smaller than that of conscientiousness. We found no significant demographic effects on wave nonresponse in one of the studies focused on aging and well-being. However, in the study primarily focused on COVID-19-related topics conducted during the pandemic, we found a few significant demographic effects. Collectively, our findings suggest that personality traits may play a more significant role than conventional demographic and household variables in predicting nonresponse patterns in high frequency (at least one survey per month) online surveys.