Measures of Association in Medicine and Epidemiology

Abstract

The measures of association evaluate the direction and magnitude of the relationship between two variables. The “effect size” of the association is described with a variety of coefficients when variables are expressed as continuous data (e.g. Pearson correlation coefficient), while if the association refers to categorical variables, the strength of the relationship between these variables is measured through difference in proportions, relative risk, odds ratio. Tests of significance evaluate statistically whether the relationship between the variables is significant or not. (Epidemiology_statistics)