112 BUSINESS APPLICATIONS OF MULTIPLE REGRESSION
for the overall model to be significant although none of the individual
slope coefficients is significant as we begin to work with the model. This
can only happen when the model has a great deal of multicollinearity.
When this happens, it will always be the case that the following pro-
cedures will result in at least one variable becoming significant. Stated
another way, when the overall model is significant, we must end up with
at least one significant slope coefficient, regardless of how they appear in
the original results.
We call the process of moving from including every variable to only
including those variables that provide statistical value model building. In
business, we use ...