Linear Correlation and Regression

Abstract

The main objective in most epidemiologic studies is the measurement of the association between exposure and outcome. Chapter 13 focused on testing the difference in outcome measured on a continuous scale in two (or more) exposure groups. The difference between the group means reflects the extent of association between the continuous outcome and the categorical exposure, and the corresponding P value derived from a t- or z-test indicates the probability of obtaining the observed or stronger degree of association by chance under the null hypothesis. In Chapter 14, both outcome and exposure were categorical (often dichotomous). The usual measures of association between a dichotomous exposure and a dichotomous outcome are the difference in proportions in those subjects with and without the outcome in the two exposure groups and the relative risk (RR) for cohort studies and the odds ratio (OR) for case-control studies. These measures are then usually tested for statistical significance using the χ² test or Fisher exact test.