Intermediate coding versus direct mapping accounts for the SNARC effect: Santens and Gevers (2008) revisited.

Intermediate coding accounts have been used to provide an explanation for the spatial-numerical association of response codes (SNARC) effect and can be contrasted with the classic direct spatial mapping account of such an association (i.e., from the mental number line directly to the responses). Importantly, a study by Santens and Gevers (2008)'s has been widely referenced in support of intermediate coding, with 90 citations at the time of writing. Hence, the current study attempted to replicate Santens and Gevers (2008)'s findings using identical methodology and a larger sample size. As in that study, participants engaged in a single-digit, magnitude judgment task. Smaller and larger responses were made using keys that were either close to or far from a middle start key with half of the participants responding to the left of the middle key and half responding to the right. Results indicate that close responses were associated with smaller numbers and far responses associated with larger numbers. However, unlike in Santens and Gevers (2008), this association could only be observed for the group of participants who responded to the right of the middle key. As will be discussed, although the presence of such an association effect for right-side responders is consistent with both the intermediate coding and direct mapping accounts, the lack of such an association effect for left-side responders is not completely consistent with either account.