Short-term retention of spatial information.

Four experiments investigated the recall of a subspan set of spatial locations over short intervals of 5 and 15 s. The intervals were filled by one of three activities: simple tapping, repeated tapping at spatial targets and backwards counting. Spatial tapping, which decreases spatial memory span, led to further small but significant errors in recall, as did backward counting. These errors were larger than those found with simple tapping over the same intervals. Backwards counting led to further decreases in recall performance if it was also present during encoding, but this was not the case for spatial tapping. Spatial tapping had little effect on recall of three digits, whereas backwards counting had a large effect, which was much larger than that found with spatial memory items in any condition. The results are interpreted in terms of the use of place-keeping functions in spatial memory sequences, which may not be specific to spatial material.