gave larger proportion of appropriate choices. The differences between the groups, although significant, were relatively small, suggesting that the major effect of number of pairings may occur at some very low number, and that by the time we reach 25 or 50 pairings, the effects are essentially complete