Miller, W. S. (1924). The Variation and Significance of Intelligence Quotients Obtained from Group Tests. Journal of Educational Psychology, 15(6), 359-366.
http://dx.doi.org/10.1037/h0074923
That certain group tests have shown a variation in intelligence quotients has been demonstrated. This, then, sets a problem for those who use group tests as indices to equate the IQ's of the different tests. Here is taken into consideration the advisability of equating IQ's from 9 group tests and the problem of test validation. The material is composed of the Miller Mental Ability Form A, Haggerty Intelligence Examination Delta 2 (given June 10, 1922); Army Alpha Form 8, Illinois General Intelligence Scale Form 1, Terman Group Test of Mental Ability Form A (given April 9, 1923); Dearborn Group Test of Intelligence Series II-C (given April 10, 1923); Otis Self-Administering Higher Examination Form A (given May 12, 1923); Stanford-Binet Individual Test (given between March 1 and May 15, 1923). The results show the mental age norms to vary so much that it is impossible to interpret the IQ's from all group tests according to the S-B standard. By translating the IQ's into variability units (tenths of S.D. with zero at 5 S.D. negative) the direct comparison of some group tests by means of the IQ is greatly facilitated. From Psych Bulletin 22:05:00509. (PsycINFO Database Record (c) 2016 APA, all rights reserved)