ABSTRACT
The great methodologists from psychology, Jacob and Patricia Cohen, entitled their textbook Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences (1975); this is an old edition, but it happens to be my favorite. In titling their book this way, they emphasized that correlation and linear regression are essentially the same thing. I have followed their thinking and have introduced the correlation coefficient here in the context of linear regression. In statistical parlance, that simple little straight line is a linear regression. What we did was regress locus of control on cultural consonance. The underlying model is the straight line, and the measure of the goodness of fit of the straight line to the data is the correlation coefficient.
You will run into lots of discussions of the correlation coefficient and lots of uses of the correlation coefficient that have no reference whatsoever to 92the underlying linear regression model. That’s fine, because the correlation coefficient is a wonderfully flexible tool that we can put to use in a whole variety of ways. At its foundation, however, is the linear regression model. In fact, in much (really most) of statistical modeling we think about the world in terms of straight lines (and more about that later).
