THE PROBLEM IS INTRODUCED IN ITS HISTORICAL BACKGROUND. THE SUBSTRATA-FACTOR THEORY IS PRESENTED WITH ITS STATISTICAL MODEL, ITS ASSUMPTIONS, AND THEIR RATIONALE. A CRITICISM OF THE LAST STEP IN SUBSTRATA ANALYSIS IS ENTERTAINED, REJECTED AS IRRELEVANT, AND, THEREFORE, INVALID BECAUSE EXPLANATIONS OF THE CAUSE OR CAUSES OF A CORRELATION MUST FOLLOW NATURALLY FROM THE BASIC ASSUMPTIONS OF THE THEORETICAL MODEL WHEREIN IT IS USED. THE THEORY ASSUMES, IN KEEPING WITH THE MATHEMATICAL DEFINITION, THAT A CORRELATION EXPRESSES A MUTUAL RELATIONSHIP BETWEEN INTACT VARIABLES AND NEED NOT APPEAL TO THE OVERLAPPING OF IDENTICAL AND INDEPENDENT ELEMENTS -- THE BASIC ASSUMPTION UNDERLYING THE CRITICISM. ASSUMPTIONS ARE SUPPORTED BY MATHEMATICAL, PSYCHOMETRIC, AND NEUROLOGICAL EVIDENCE. A BIBLIOGRAPHY IS INCLUDED. THIS ARTICLE APPEARED IN "READING RESEARCH QUARTERLY," VOLUME 1, NUMBER 1, FALL 1965. (AUTHOR)