Nonresponse at three stages in personality research: Insights based on (Danish) register data of a representative potential participant pool

Study participants are vital to personality research. Thus, individuals not following study invitations (self-selection), not completing all administered measures (showing dropout) and/or not participating at subsequent measurement occasions (showing attrition) pose a significant challenge to personality research. Building on existing research documenting that non-response does not occur randomly, we herein test for potential predictors of non-responding at each of the three stages by combining official register (census) data from a randomly selected participant pool of 100,175 adults, representative for the target population (adults living in Denmark), with data from a personality panel (N = 14,071) recruited from this participant pool. We find that each socio-demographic characteristic considered (i.e., sex, age, marital status, migration background, income, highest completed education, and region of residence) as well as five of the six HEXACO personality dimensions (Honesty-Humility, Emotionality, Extraversion, Conscientiousness, and Openness to Experience; but not Agreeableness vs. Anger) were associated with self-selection, selective dropout, and/or selective attrition. We further compare the sample representativeness at each of the stages and illustrate how to assess potential biases (regarding the obtained HEXACO scores) due to self-selection.