A cross-sectional comparative study on the differences in personality traits and demographic associations among samples from different survey locations

This study compared personality trait scores between university student samples surveyed in Shanghai and Xinjiang and examined their associations with selected demographic characteristics. A cross-sectional questionnaire survey was conducted among 1168 university students in Shanghai and Xinjiang. Participants completed the Chinese version of the Eysenck Personality Questionnaire–Revised Short Scale. Univariate comparisons were performed using Mann–Whitney U tests with Holm–Bonferroni adjustment. Age-adjusted Type III ANCOVA models were used as the primary analyses. All main effects, two-way interactions, three-way interactions, and covariates were retained and reported irrespective of statistical significance. Benjamini–Hochberg false-discovery-rate correction was applied across the terms within each model set. The corresponding complete unadjusted Type III ANOVA models were treated as sensitivity analyses and reported separately. For Neuroticism, the standardized Lie scale score was additionally included as a covariate. None of the univariate comparisons remained statistically significant after Holm–Bonferroni correction. In the primary ANCOVA models, no demographic main effect or interaction remained statistically significant after false-discovery-rate correction for Extraversion, Neuroticism, or the Lie scale. The standardized Lie scale covariate was strongly associated with Neuroticism in all three model sets. A small number of survey-site, gender, and age terms remained significant for Psychoticism. Evidence for robust demographic differences in Extraversion, Neuroticism, and the Lie scale was limited after covariate adjustment and multiple-comparison correction. The findings should be interpreted as exploratory associations within the two surveyed samples rather than as evidence of broader regional or ethnic-group differences. Further multi-institutional studies with more balanced samples and broader socioeconomic and educational measurements are needed.