Tables to Facilitate the Computation of Coefficients of Correlation by Rank Differences Method

Abstract

Tables are given to facilitate the computation of the coefficient of correlation by the rank difference method. The computer first arranges his two series of measurements to be correlated in rank order. The difference between the two ranks is obtained and then the square of these differences computed. The table presented in this article gives the square of the differences from 0 to 80 by halves, i.e., square of .5, 1.0, 1.5, etc., up to 79.5. These squares of differences are added giving Σ D 2 . This sum, Σ D 2 is found in the body of the second table of this article. The head of each column indicates the number of cases. Thus if 21 cases are ranked, one runs down the column headed 21 to find the entry nearest the obtained Σ D 2 and then reads the entry on that horizontal line which is under the heading ∠?← This entry, ∠?→ is the coefficient of correlation. Reprints of this article with the tables may be obtained from the Journal of Applied Psychology. Orders should be addressed to Miss Florence Chandler, Clark University, Worcester, Mass. From Psych Bulletin 18:01:00005. (PsycINFO Database Record (c) 2016 APA, all rights reserved)