§ Abstract
Visual cognition theories propose that mental imagery and perceptual processing share neuralmechanisms, predicting that individuals with stronger imagery abilities should exhibit enhancedtemporal integration across diverse visual tasks. We tested this hypothesis using computationalmodeling to examine whether visual imagery vividness modulates serial dependence in visualworking memory and sequential effects in mental rotation. Participants completed visual work-ing memory and mental rotation tasks alongside imagery vividness questionnaires. We appliedderivative-of-Gaussian modeling to characterize how previous trial information influencescurrent performance, extracting individual difference parameters for bias strength (amplitude),range of influence (width), and the balance between perceptual stability and change detection(zero-crossing). Despite rigorous computational approaches and adequate statistical power,visual imagery vividness failed to predict any aspect of temporal integration across both tasks.Analyses based on over 170 participants per task showed that those with stronger self-reportedimagery exhibited neither enhanced serial dependence in working memory nor stronger se-quential facilitation in mental rotation. Cross-task correlations between temporal integrationmechanisms were absent, precluding a planned analysis of whether imagery strength acted as amoderating factor. These null findings challenge prevalent theories proposing shared neuralsubstrates between imagery and perception, suggesting that subjective imagery experiencesmay not meaningfully predict performance on fundamental cognitive tasks. The results in-dicate that individual differences in visual cognition operate through more domain-specificmechanisms than previously assumed, with important implications for imagery-based training programs and computational models of visual processing that incorporate individual variability