45 retarded Ss were required to learn serial word lists of three lengths (4, 8, and 12 words). The percentages of errors in the middle half of each list length were compared to test the McCrary-Hunter invariance hypothesis, with list length as the dimension of task difficulty. An analysis of variance revealed a list-length effect that approached, but did not attain, the .05 level of significance. Orthogonal comparisons of the treatment sums, however, revealed a significant linear regression (P < .05) of error percentages in the middle position as a function of list length. The hypothesis of invariance in relative serial position error was not supported for different lengths of list.
Get full access to this article
View all access options for this article.
References
1.
GirardeauF. L.EllisN. R.Rote verbal learning by normal and retarded children. Amer. J. ment. Defic., 1964, 68, 525–532.
2.
McCraryJ. W.HunterW. S.Serial position curves in verbal learning. Science, 1953, 117, 131–134.
3.
McManisD. L.Relative errors with serial lists of different lengths. Amer. J. ment. Defic., 1965, in press.
4.
WetherfordJ. W.Relative serial position effects as a parameter of verbal learning: a test of the McCrary-Hunter hypothesis. Dissert. Abstr., 1962, 23, 3, 1083–1084.
