Abstract: Previous behavioral research has shown evidence for the dissociation between the remembering of an event and the contextual details in which this event occurred. The first task is called item memory and the second is referred to as source memory. The dissociation between these two tasks is further reinforced by event-related potentials (ERP) studies. The binding of distinct sources with items also differs under certain circumstances, which has been confirmed by using ERP or other measures. Moreover, the testing style, aging, material and many other factors can regulate the spatial and temporal distributive characteristic of the old/new effect for both item memory and source memory. To verify whether the minimum impact of testing paradigms on the neural correlation of color source retrieval reported by Cycowicz and his collaborators also held true in other sources, two ERP experiments were conducted to investigate the impact of two types of testing paradigms (exclusion and three-key paradigms) on the temporal and the spatial distributions of the old/new effect for the location source retrieval task, in which both of these experiments used pictures as stimuli.
Thirty-three subjects participated in the two experiments, sixteen of them for the first experiment, and the rest for the second one. In the first experiment, stimuli were presented either on the left of the screen or on the right side during the study phase and then two tests were performed separately with pictures presented on the center of the screen. One test task was to discriminate the old items from the new ones, and the other task was to judge the items presented on certain location during the study phase as targets and to judge other ones as non-targets. In the second experiment, the source test was to judge each item with three different keys based on its context during the study phase. Two-way repeated-measures ANOVAs were performed with response (2 levels) and electrode (4 levels) served as within-subject variables to test the old/new effect associated with each retrieval task and the Greenhouse-Geisser epsilon (ε) correction was used where appropriate.
For the first experiment, the spatial distribution of the positive-going old/new effect elicited by the location source retrieval task was wider than that of the item recognition task. Whereas, the second experiment recorded reliable negative-going old/new effect correlated with the location source retrieval task relative to that of the item recognition task.
Unlike the findings reported by Cycowicz and his collaborators, the results from the current experiments showed that the spatial and temporal distributions of the old/new effects for the location source retrieval tasks with different testing paradigms were also distinct, which indicated the fact that the neural correlation of this task was regulated by the testing style; and the possible reason for this difference was that the binding of location source with item differed from that of the color source. The current results, together with those from Cycowicz and his collaborators’, suggest that the conclusion based on color source studies holds true only in limited ranges of episodic sources.

Key words: source location, item recognition, testing paradigm, old/new effect, LPN