Laboratory research has established that face recognition memory performance for ownrace faces is better than for other-race faces. Three studies are reported exploring the possibility that the other-race effect will generalize to voice recognition memory. Recognition memory performance for non-native American speakers speaking both English and their native languages was compared with memory for native American speakers. With relatively long speech samples, accented voices were no more difficult to recognize than were unaccented voices; reducing the speech sample duration decreased recognition memory for accented and unaccented voices, but the reduction was greater for accented voices.

Three experiments attempted to explain why high clusterers recall more than low clusterers. In Experiment 1, we cued high and low clusterers to recall category exemplars contiguously, noncontiguously, or by free recall. Performance was best contiguously, worst noncontiguously, and intermediate in free recall. Surprisingly, the magnitude of the high-clusterer recall superiority was the same in each recall condition. In Experiment 2, we investigated whether differences between high and low clusterers involve differences in how lists are studied. Again, the high-low clusterer recall difference was the same when category exemplars were presented contiguously and randomly for study. Experiment 3 showed a high-clusterer recall superiority in unrelatedword lists. We conclude that the high-clusterer recall advantage is due not to clustering per se. Clustering can be viewed as a moderator variable that facilitates recall within the boundaries set by ability differences among subjects.

No abstract

Male and female subjects provided ratings of personality traits for students pictured in a large set of facial photographs. Each photograph was rated on each of eight traits. Trait ratings were also provided for subjects' impressions of "typical" male and female college students. Male and female subjects did not differ in their ratings of photographs or of typical students. Sex of the picture (and of the typical student) did reliably influence the ratings for several traits. Certain groups of traits appeared to form intercorrelated clusters, suggesting the operation of a halo effect.

For the purpose of specifying-the relationship between item strength and input position, six study lists were presented for immediate free recall, followed by a relearning list composed of words taken from beginning, middle, or end positions of the study lists. Better relearning performance was obtained for beginning and middle items, relative to end items, for words previously recalled in study. For words not previously recalled, end-item performance was better in relearning. A hypothesis of selective attention, wherein subjects attend to subjectively easy items in prerecency, but not recency, positions, is advanced to explain the results.