Contextual associations and memory for serial position.

Three experiments used a combination list-discrimination and position-judgment task to investigate the role of contextual factors in incidental memory for serial position. In Experiment I, two temporally denned lists were presented; in Experiments II and III, there were four and three lists, respectively. Following presentation of the lists, 5s made judgments of the list membership and within-list position of test words. Judgment frequencies revealed: (a) a temporal factor, affecting list identification and producing strong primacy and recency effects; and (b) an effect of position, when more than two lists were used, such that a word assigned to an incorrect list still tended to be placed in the correct within-list position. When the retention interval was lengthened the effects of primacy and within-list position were unaffected, while the effect of recency was reduced. An interpretation is offered which assumes judgments of serial position are based on contextual associations. If S is shown a list of words, and one of them is later presented as a probe on a memory test, S is not only able to say that the word occurred in the list, but in addi-tion he knows something about the serial position the word occupied (Hintzman & Block, 1971; Zimmerman & Underwood, 1968). This ability is prima facie evidence for the existence of some sort of "time tag." The time tag is of general theoretical interest because it could play an important role in many memory tasks: In recognition memory S must know that an item was presented during the experiment, in short-term retention that it occurred on the cur-rent trial, and in a variety of other tasks that it occurred in a particular temporally defined list. The possible mediating func-1 The research reported herein was performed