A(B)OVER-BAR WITH MULTIPLE WELLS .1. WHY ARE MULTIPLE WELLS SOMETIMES EASIER THAN 2 WELLS .2. MEMORY OR MEMORY PLUS INHIBITION

Seventy-two infants of 91/2-10 months were tested with a 7-well apparatus and a 5-s delay. Other researchers had found worse performance with 2 wells than with more wells, leading some to question the role of memory in ABBAR performance. We hypothesized that a difference in procedure might have caused the performance difference: In 2-well studies, both wells are covered simultaneously. In multiple well studies, only the correct well is uncovered and re-covered, which might help one maintain attention on the correct well. This was tested in Part 1 with 3 conditions: (a) all wells covered simultaneously with slits, (b) toy hidden by lowering it through a slit (last action at correct well), and (c) uncovering correct well only, hiding toy, re-covering that well (again, last action at correct well). Predictions were (a) worse performance in Condition 1 than in Conditions 2 or 3 or 2-well ABBAR studies and (b) comparable performance in Conditions 2 and 3, equal or better than in 2-well ABBAR studies. Predictions were confirmed. The use of 7 wells enabled the testing, in Part 2, of whether errors are due to forgetting or to forgetting plus a failure to inhibit a rewarded response. If the problem is memory, errors should be normally distributed about the correct well. If the problem is memory + inhibition, more errors should occur in the direction of the previously correct well. These hypotheses were tested using 2 wells between A and B and 2 wells on the other side of B. Errors occurred disproportionately in the direction of the previously correct well, suggesting that ABBAR requires both memory and inhibition.