U-shaped engagement patterns in online mental health surveys: response times, careless responding, and demographic influences

This study investigated response patterns and data quality in online psychological surveys through analysis of a large-scale mental health screening dataset ( N = 24,292) from Chinese university students. Using detailed response time data and established quality indicators, we examined temporal patterns of survey engagement and identified factors associated with careless responding. Results revealed systematic U-shaped distributions of response times across psychological instruments, with higher engagement at the beginning and end of scales, and notably lower engagement in middle sections. Careless response rates varied among different psychological instruments (37.02-52.77%), with scale position potentially contributing to these differences in the fixed-order battery. Cluster analysis identified four distinct respondent profiles based on educational and substance use characteristics. Counterintuitively, higher education levels were associated with increased careless responding (57.43%), while moderate social behaviors correlated with higher response quality. The study provides empirical evidence for temporal dynamics in online survey engagement and demonstrates complex interactions between participant characteristics, survey structure, and response quality. These findings have important implications for online survey methodology and data quality control in psychological assessment, particularly regarding the need for dynamic monitoring mechanisms and strategic survey design to maintain participant engagement.