Abstract

When responding to numerals with left-right key presses, performance is better for pairings of small numbers to left responses and large numbers to right responses than for the opposite pairings. Two accounts have been proposed to explain this Spatial Numerical Association of Response Code (SNARC) effect: the horizontal number line account which ascribes the SNARC effect to numbers coded as left or right and the polarity correspondence account which attributes it to the magnitude information being coded as a positive or negative polarity. This study examined whether the SNARC effect is due to the spatial correspondence between the number location on the number line and the response location, or to the correspondence between the polarity codes of the number magnitudes and response locations. When participants responded to the magnitude of an Arabic numeral presented at the left or right to fixation in Experiment 1, the SNARC effect was constant regardless of spatial correspondence between the stimulus and response locations. In contrast, when the numeral was presented above or below fixation in Experiment 2, the SNARC effect was smaller for the up-left/down-right pairings than for the up-right/down-left pairings. These results support a view that polarity correspondence contributes to the SNARC effect.