An evaluation of measures used in studying temporal codes for words within a list

Abstract

Experiment I evaluated three response measures which have been used in the study of temporal codes for memories. The three measures were judgments of recency, position, and lag. Earlier studies had suggested that subjects could make valid position judgments without being able to make valid lag judgments. This finding was confirmed; although the subjects learned a great deal about temporal ordering when asked to use lag judgments, the lag judgments per se were very poorly related to true lag. Transfer from lag judgments to recency judgments and from position judgments to recency judgments was essentially 100%. Transfer from recency judgments and lag judgments to position judgments was partial. It appeared that recency judgments can be mediated with little knowledge of the absolute position in the study list held by one or both of the items in the recency test pairs. Experiment II demonstrated that, under certain circumstances, lag judgments could be valid. The results raised the possibility that when lag judgments are not valid the magnitude of the lag judgments may be influenced by certain contrast effects. The data also suggested that serial learning can be directly implicated in recency judgments, hence, in temporal coding.