Tailoring perception and action to the task at hand

Abstract

Two experiments investigated whether and how task demands affect the way perceptual events and actions are cognitively represented, and how this affects performance. Subjects performed two-dimensional Simon tasks that alternated with, or were embedded into, a logically unrelated “priming task”, in which the relevance of horizontal and vertical dimensions varied frequently. Making the horizontal dimension relevant in the priming task increased the horizontal Simon effect, and making the vertical dimension relevant increased the vertical Simon effect. These findings suggest that stimulus and response representations are not invariant but are tailored to the current task demands. This has implications for models of perception and action planning, stimulus–response compatibility, and executive control.

Notes

¹ For the sake of completeness, we also analysed the data from the priming task. Missing and anticipatory responses were excluded and mean RTs and PEs were analysed as a function of primed spatial dimension, the spatial compatibility between response and the stimulus location on the unprimed dimension, and of horizontal and vertical compatibility in the (preceding) Simon task. Reliable main effects were obtained for compatibility in the priming task, F(1, 22) = 24.26, p<.001, MSE = 19,491.58, and of horizontal compatibility, F(1, 22) = 5.15, p<.05, MSE = 10,000.31, and vertical compatibility, F(1, 22) = 13.85, p<.001, MSE = 6525.14, in the Simon task. All these effects entered into a three-way interaction, F(1, 22) = 7.07, p<.05, MSE = 2908.91, and vertical compatibility was further modulated by priming, F(1, 22) = 7.79, p<.05, MSE = 11,150.46. The latter interaction was due to the fact that performance in vertical priming conditions was strongly affected by vertical compatibility in the Simon task (711 vs. 774 ms), whereas performance in horizontal priming conditions was not (759 vs. 759 ms). The three-way interaction reflected that performance on the compatibility effect in the priming task was increased after fully compatible (85 ms) and fully incompatible (89 ms) Simon trials as compared to Simon trials of mixed compatibility trials (63 and 50 ms for vertically compatible/horizontally incompatible and vertically incompatible/horizontally compatible, trials respectively). PEs showed a main effect of compatibility on the unprimed dimension in the priming task, F(1, 22) = 30.54, p<.001, MSE = 208.84, and an interaction of vertical and horizontal compatibility in the Simon task, F(1, 22) = 7.99, p<.01, MSE = 49.12, that was further modified by priming, F(1, 22) = 13.98, p<.01, MSE = 32.88. A four-way interaction, F(1, 22) = 17.85, p<.001, MSE = 32.77, indicated that incompatible priming trials in the vertically primed dimension were modified by compatibility in the Simon trial.

² For the sake of completeness, we also analysed the data from the priming task. Missing and anticipatory responses were excluded and mean RTs and PEs were analysed as a function of primed spatial dimension, the spatial compatibility between response and the stimulus location on the unprimed dimension, and of horizontal and vertical compatibility in the (preceding) Simon task. Reliable main effects were obtained for compatibility in the priming task, F(1, 22) = 24.26, p<.001, MSE = 19,491.58, and of horizontal compatibility, F(1, 22) = 5.15, p<.05, MSE = 10,000.31, and vertical compatibility, F(1, 22) = 13.85, p<.001, MSE = 6525.14, in the Simon task. All these effects entered into a three-way interaction, F(1, 22) = 7.07, p<.05, MSE = 2908.91, and vertical compatibility was further modulated by priming, F(1, 22) = 7.79, p<.05, MSE = 11,150.46. The latter interaction was due to the fact that performance in vertical priming conditions was strongly affected by vertical compatibility in the Simon task (711 vs. 774 ms), whereas performance in horizontal priming conditions was not (759 vs. 759 ms). The three-way interaction reflected that performance on the compatibility effect in the priming task was increased after fully compatible (85 ms) and fully incompatible (89 ms) Simon trials as compared to Simon trials of mixed compatibility trials (63 and 50 ms for vertically compatible/horizontally incompatible and vertically incompatible/horizontally compatible, trials respectively). PEs showed a main effect of compatibility on the unprimed dimension in the priming task, F(1, 22) = 30.54, p<.001, MSE = 208.84, and an interaction of vertical and horizontal compatibility in the Simon task, F(1, 22) = 7.99, p<.01, MSE = 49.12, that was further modified by priming, F(1, 22) = 13.98, p<.01, MSE = 32.88. A four-way interaction, F(1, 22) = 17.85, p<.001, MSE = 32.77, indicated that incompatible priming trials in the vertically primed dimension were modified by compatibility in the Simon trial.