Distinguishing familiarity from list search responses in a reaction time task

This experiment tested the claim of the model proposed by Atkinson and Juola (in press) that recognition responses in an RT task are a mixture of fast responses on the basis of the item’ s familiarity and slower responses based on a list search for the item. All Ss had to learn two 16-word lists (A and B). Then they had to respond alternately to “Is the following a List A word? ” and “Is the following a List B word? ” In the control task, the members of one list served as distractors for the other. Hence, the familiarity values of the targets and distractors were equal, and the only basis for response was a list search. In the experimental task, the distractors were not list words. Hence their familiarity was lower, and familiarity responses could be made. There were no differences between two groups of Ss when they all responded in the control task. But when half the Ss were switched to the experimental task, they responded significantly faster than the Ss who continued in the control task. In the experimental task, distractors were rejected quicker than targets were identified, but in the control task, the reverse was true. These results support the hypothesis that the control task required a list search while a familiarity judgment was possible in the experimental task.