An Empirical Comparison of Pupil Achievement and Other Variables in Graded and Ungraded Classes

During the past decade, interest in classroom organization and structure has increased, particularly interest in the ungraded, or nongraded, school. Goodlad and Anderson (1959) listed approximately fifty communities in which schools were using this type of organization, usually at the primary level. Much has been written regarding the purported superiority of the nongraded system (Anderson, 1962; Downs, 1960; Goodlad and Anderson, 1959; Gore, 1962; Time, 1963). Increased academic achievement is frequently mentioned. However, surprisingly little empirical research on this point has been reported. Regarding the comparative achievement of pupils in graded and ungraded schools, Goodlad and Anderson (1959, p. 57) said: ... there is no evidence to suggest anything. We have little more than inadequate firsthand impressions to go on. Later, in surveying the research on classroom organization, Goodlad (1960, p. 222) again commented, Nongrading is supported by some plausible-sounding claims and theories rather than by research. Results of the studies that have been done appear perplexingly contradictory until the research techniques are scrutinized. Skapski (1960), Ingram (1960), and Hillson, Jones, and Moore (1964) reported that the ungraded program produced significantly superior achievement. Carbone (1961) had the opposite result, and Kierstead (1963) found no significant difference. Serious methodological problems centering around noncomparable treatment groups attenuated the results of Skapski's and Ingram's studies. Carbone's design, on the other hand, may not have been completely fair to the ungraded pupils, since the comparison was made only after they had been channeled back into graded classes. The experimental design used by Hillson, Jones, and Moore appears to be the best to date. However, the small size of their samples (pupils, combined N of 52;