Search

The influence of a stimulus suffix was examined in yes-no recognition and probe-recall tasks. As with previous findings with serial recall, reliable suffix effects were obtained in probe recall. However, no suffix effects were found in the recognition task. Both serial and probe-recall tasks require the retention of order and item information, while only item information is required in recognition. An interpretation of the suffix effect based on interference with order information is presented.

Comparisons were made between groups of old and young subjects in a dichotic task. In the first experiment old and young subjects had digit spans representative of their respective populations, and in accord with previous dichotic research younger subjects recalled with greater accuracy. The main effect of group and the Group x Test (ear reported first vs ear reported second) interaction were reliable with both serial and free recall scoring. In the second experiment young subjects were matched to old subjects on the basis of digit span. Matching eliminated the group and Group x Test interaction with seriall recall and the group effect with free recall, suggesting that digit span and dichotic memory are mediated by common mechanisms. Error trends and the Group x Test interactions with free recall revealed that some age related differences persisted even after matching. These differences are discussed in terms of regression artifact and processing capacity.

Retention of old (n = 27; M age = 72.4 years) and young (n = 46; age = 20.5 years) adults was compared in a Brown-Peterson task. Participants were instructed to recall letters after either reading digit pairs or summing each pair of digits and reporting the sum as old or even. Old adults recalled a substantially smaller proportion of letters correctly than did the young adults, and the magnitude of the difference between groups was related positively to the length of the retention interval. The main effect of age and the age by retention interval interaction were significant with both serial and free recall scoring criteria. In a multiple regression analysis with letter span, age and interpolated task regressed on Brown-Peterson recall, letter span, and interpolated task accounted for most of the variance in performance.