Woodworth, R. S. (1939). Combining the results of several tests: A study in statistical method. In R. S. Woodworth, Psychological issues: Selected papers of Robert S. Woodworth, professor of psychology, Columbia University (pp. 257-283). New York, NY, US: Columbia University Press.
http://dx.doi.org/10.1037/13498-019
This reprinted article originally appeared in Psychological Review, 1912, 19(2), 97-123. (The following abstract of the original article appeared in record 1926-03199-001.) Investigates the use of a statistical method which gives each individual's average standing in a number of tests. Attempts to work out simplified formulae which can be used for computing correlations. The assumption underlying the method is - the average and the measure of variability mean the same thing in different distributions. The method involves reducing the original measures to terms of the a.d.(mean square deviation) for finding (1) the Pearson correlation coefficient (2) the average correlation and (3) the Spearman correction for attenuation. The method is less laborious, and gives insight into the causes of correlation, in terms of individual contribution to the average result. (PsycINFO Database Record (c) 2016 APA, all rights reserved)