ABSTRACT
Having set the stage by describing how mental health levels dif fered among Midtowners in different levels of social and biological age, gender, marital status, and Parental SES, we proceed to the use of correlations and adjusted correlations (path coefficients, which we also call predictions) to test the propositions of MHIM2. We are constrained by considerations of space and expense from presenting the many tables and figures needed to statistically describe a complicated study such as MHIM2. But this presents the opportunity to discursively present the methods and findings in prose rather than numbers. In so doing, we believe that we shall be making comments which reflect what a statis tically knowledgeable reader would make of the correlations and path coefficients.1
