ABSTRACT
Correlated variables are variables which change together in a systematic fashion, and a correlation coefficient is a measure of the degree and direction of that relationship. The mechanics of computing correlations and regression lines is an easy matter: even the simplest of hand calculators will allow quick computation of correlations and parameters of regression lines. Correlations are independent of the units in which measurements are taken. In fact the measurements themselves need not even be in the same units in order to obtain data and compute correlations. Regression and correlation are virtually interchangeable terms in statistics. Regression analysis is necessitated by the fact that correlations are virtually always less than +1.0 or -1.0. Regression lines are determined by using the least squares technique to result in what is called the least squares or best fit regression line. Computation of these lines can be done on almost all hand calculators and personal computers, and the mechanics are described in numerous sources.
