Global-local tendency task
Description
Understanding individual differences in global-local visuospatial processing is crucial across psychological disciplines. However, attempts to measure these differences face two fundamental challenges. Conceptually, traditional paradigms like the Navon task conflate spontaneous perceptual preference with top-down executive control, thereby potentially conflating processing ability with default tendencies . Psychometrically, these tasks frequently suffer from poor test-retest reliability, -a manifestation of the "reliability paradox,” severely limiting their utility for trait-level assessment. To address these theoretical and methodological gaps, the present study introduces and validates the Global-Local Tendency Task (GLTT), a novel, instruction-free paradigm designed to reduce the influence of cognitive control [EK1.1]on estimates of default perceptual bias.. 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.