Effects on Talker Variability in self-Paced Serial Recall

Author: Stephen D. Goldinger

Abstract:
Goldinger, Pisoni, and Logan (1991) examined serial recall of several types of spoken word lists; words within a list were produced by either a single talker or by multiple talkers, and lists consisted of either "easy" of "hard" words, as determined by a metric of word frequency and similarity neighborhood characteristics. Goldinger et al. manipulated presentation rate and found that talker variability interacted with presentation rate but word confusability did not. We concluded that talker variability affected the rehearsal processes involved in serial recall of spoken word lists. The present experiment complemented the earlier experiment by examining recall of the same spoken word lists using a self-paced listening procedure. In this procedure, subjects were able to control the presentation rate of successive items within a list. Both the listening time and the eventual recall were examined. The data showed that both stimulus dimensions affected rehearsal time, but rehearsal time differences were only predictive of recall differences between single- and multiple-talker lists. Rehearsal time differences were not predictive of differences between recall of lists of "easy" vs. "hard" words. The findings both confirm and qualify the earlier conclusions of Goldinger et al. (1991).