Over a long period, the relation between scatter and intelligence quotient has been a topic for investigation. 8 It continues to claim attention. McNemar 8 has reported upon the relationship as found with the 1937 revision of the Stanford-Binet. But the group-intelligence-test counterpart of scatter has not been frequently studied. Woodrow 4 reported an extensive analysis using the results of the Arthur and Woodrow group intelligence scale, and holding mental age constant. He refers to prior studies by Maud A. Merrill, 6 J. C. DeVoss, 6 and A. W. Brown, 7 but the Merrill study was concerned with Binet scatter, De Voss studied achievement test scores, and Brown held chronological age constant, not mental age.An examination of the Woodrow data reveals a need for further study. After reporting results at eighteen half-year mental-age levels from 8:0 to 16:11, Woodrow averaged his findings and reported that unevenness is at a minimum for average pupils, and that it is greater for both brighter and duller pupils but not as 1 Grateful acknowledgement is made to H. M. Foulds, Jr. and Howard Stoertz for assistance with the computation; to the Yale University Committee on Bursary Appointments for making this assistance possible; and to