Posture and Attention
Loading...
Date
Authors
Advisor
Smilek, Daniel
Carriere, Jonathan
Carriere, Jonathan
Journal Title
Journal ISSN
Volume Title
Publisher
University of Waterloo
Abstract
Recent research found that posture (sitting versus standing) influences performance on cognitive paradigms including the Stroop task, task-switching, and visual search (Rosenbaum et al., 2017;2018; Smith et al., 2019). The proposed mechanism suggests standing increases ‘load’, ultimately enhancing selective attention (Rosenbaum et al., 2017). Yet, early findings were ambiguous and the theory underspecified. This dissertation presents an account of systematic replications that test the robustness of the postural effect on attention, as well as includes a computational investigation into the mechanism underlying this theoretical account. Chapter 2 assessed the reliability of Rosenbaum and colleagues’ original postural effect. Chapter 3 directly replicates Rosenbaum and colleagues’ and Smith and colleagues’ Stroop experiments. Chapter 4 investigated the influence of posture on task-switching (switch-costs) and visual search (search-rates) via replications of Smith and colleagues’ paradigms. Chapter 5 applied Rosenbaum and colleagues’ theory to a computational model of attention in the Stroop task, exploring how ‘load’ and cognitive resources influence performance. The empirical experiments (Chapters 2-4) failed to produce robust postural interactions, suggesting no meaningful posture effect on Stroop, task-switching and visual search performance. Model simulations (Chapter 5) partially align with Rosenbaum and colleagues’ findings but revealed a speed-accuracy trade-off; increased load produced faster but less accurate responding rather than enhanced attention. This dissertation provides converging evidence that postural influences on cognition are not as robust as initially reported. The results underscore the importance of replication and cross-disciplinary research in establishing reliable effects and suggest caution in accepting claims without rigorous verification.