Global-local tendency task

Across two sessions, 205 participants completed the GLTT and a standard Navon-like Arrow Task (NAT). Using Hierarchical Bayesian Modeling (HBM), we estimate individual bias scores while accounting for trial-level noise. Results revealed that the GLTT's primary index of spontaneous tendency exhibits excellent test-retest reliability (r = .84), substantially outperforming the conventional NAT global bias score (r = .65).

Furthermore, we found moderate, interpretable convergence between the GLTT and the NAT (r = .41), consistent with the GLTT capturing spontaneous tendency that is not fully captured by task-directed NAT performance.. These findings establish the GLTT as a psychometrically robust tool that addressesthe limitations of traditional paradigms, providing a newly validated framework for advancing individual-differences research in cognitive science and psychopathology.