Transforming variables to standard form
There are occasions when it is desirable to convert a variable into a more convenient form by standardizing it. A standardized variable is one that has been transformed so that it has a mean of zero and a variance of one. Standardization can be very useful because it enables us to compare the regression of a dependent variable on different independent variables. It is much easier to compare the relative magnitudes of different regression coefficients if the variables have the same variances. For example, it is difficult to compare the regression of income on education with the regression of income on occupational prestige because the differences in the regression coefficients for these two independent variables are attributable, at least in part, to the fact that they have different variances. By converting both of these independent variables to standard form, we can compare the change in income we expect with a one standard deviation change in education to the change in income we expect with a one standard deviation change in occupational prestige.
KeywordsPublic Health Standard Deviation Regression Analysis Regression Coefficient Absolute Deviation
Unable to display preview. Download preview PDF.