Representational context modulates the direction and transiency of practice effects on SNARC.

The Spatial-Numerical Association of Response Codes (SNARC) effect typically results in faster left-hand responses for small numbers and right-hand responses for large numbers, aligning with the concept of the mental number line (MNL). It is a robust but a flexible phenomenon that can reverse direction depending on the spatial-numerical mappings employed. This study investigates the potential modulatory effects of two contrasting representational contexts (ruler vs. clockface) on the emergence and persistence of the SNARC effect under two opposing spatial-numerical practices (MNL-compatible vs. MNL-incompatible). In Experiment 1, a magnitude classification task was employed as a practice session including either MNL-compatible or MNL-incompatible stimulus-response mappings, and the transfer and transiency of practice effects were examined by engaging participants in three test parity judgment tasks administered 5 min, 1 day, and 1 week after the practice session. In Experiment 2, different representational contexts were introduced during practice sessions. Experiment 2a utilized an image of a ruler as the context consistent with the MNL, while Experiment 2b employed a clockface image as an inconsistent context. Participants underwent testing three times to assess changes in performance over time. Results revealed that MNL-compatible practice effects did not transfer while MNL-incompatible practice effects resulted in a reverse SNARC effect persisting for up to 1 day. However, introducing the ruler context eliminated this transfer, while the clockface context reduced the persistence of the practice effect.