AN ALTERNATE CHANNEL THEORY OF THE STROOP EFFECT
Virtually all existing accounts of the Stroop Effect (SE) rely on a coactive architecture, wherein information from both the color and word channels is integrated. We challenge this assumption by presenting a new formal alternate-channel theory of the SE. The theory assumes that on each trial exactly one, i.e. color or word but not both, channel determines response. Performance (accuracy, RT) emerges from a mixture of color and word operative trials. The theory is shown to account for Stroop facilitation, interference and the total effect, with respect to both accuracy and mean RT. Nevertheless the model is limited by the strong quantitative bond between accuracy and latency: Given a low error rate, 'small' temporal effects are predicted.