Extra-List Intrusions in Immediate Free Recall as a Function of Associative Strength in Children

Abstract

After collecting word-association norms for 12 high frequency responses to 36 stimulus words drawn from the Minnesota norms (3), Deese (2) found that the relative frequency of occurrence of a stimulus word as an associate to these responses was significantly correlated with the extent to which the stimulus word occurred as an extra-list intrusion in tests of immediate free recall. Using college students as Ss, the product-moment correlation between the frequency of occurrence of the stimulus word as an associate and the frequency of occurrence of the stimulus word as an intrusion was ,873. The purpose of the present study was to examine this same relationship in grade school children. The procedure employed by Deese (2) was followed as closely as possible with three exceptions: fewer recall lists and words per list were used. the recall tests were administered in groups rather than individually (1) and the word associations were collected by having E read the stimulus words aloud instead of having the stimulus words printed on sheets of paper. Ss were school children drawn from the 4th, 5th, and 6th grades. In Part I of the experiment, 19 free recall tests consisting of 10 high frequency responses each were administered to 29 Ss each from these grades. The stimulus and response words were culled from word-association norms collected by Simon, Lotsof, and Pease (4) from 399 4th. 5th. and 6th grade children. Jn Part I1 of the experiment, word-association norms were collected from 90 different Ss drawn from the same population as that employed in Part I of the experiment. For all grades combined, the product-moment correlation between the number of times the stimulus word occurred as a free association in Part I1 and the number of times the stimulus word occurred as an intrusion in Pan I was ,725 (p< .01). The correlations for the 4th, 5th, and 6th grades, respectively, were ,675, ,719, and ,604 (p < .O1 in all three instances). None of the latter ss were significantly different from one another. These results agree well with those reported by Deese (1) and suggest that his findings may reliably be extended to include children in the 4th, 5th, and 6th grades.