CORRELATION AS A DEPENDENT VARIABLE

THE CORRELATION BETWEEN 2 DEPENDENT VARIABLES IS SUBJECT TO THE SAME EFFECTS AS ARE THE MEANS ON 1 VARIABLE. FOR EXAMPLE, THE CORRELATION BETWEEN THE 1ST AND LAST TRIALS MAY VARY WITH CONDITIONS OF PRACTICE, JUST AS THE TRIAL MEANS MAY. THE DEPENDENCE, MOREOVER, OF A CORRELATION UPON SOME EXPERIMENTAL CONDITIONS AND ITS INDEPENDENCE OF OTHERS MAY BE RELEVANT INFORMATION. IN THIS PAPER THE CHI SQUARE TEST FOR INDEPENDENT CORRELATIONS IS EXTENDED TO THE ANALYSIS OF CORRELATIONS, ACTUALLY THEIR Z TRANSFORMS, INTO MAIN, SUBSAMPLE, AND INTERACTION SOURCES OF VARIATION. CONSIDERATION IS GIVEN TO HANDLING AVERAGES OF CORRELATIONS BASED ON THE SAME SUBJECTS AND SAMPLES OF UNEQUAL SIZE. (PsycINFO Database Record (c) 2006 APA, all rights reserved). © 1968 American Psychological Association.