Can you relate in multiple ways? Multiple linear regression and stepwise regression - Improving the User Experience through Practical Data Analytics - page 257

FIGURE 10.12
SPSS output for Behemoth.com data.
(see arrow in Figure 10.11 ). We have not discussed “Method” before, and have left
the method as “Enter,” the default. In the next section, this will be an issue of note.)
The results, of course, are the same as we obtained from the Excel software,
except, as noted in previous chapters, for most everything in SPSS being rounded to
three digits. We get the same “big 7” as signiicant variables in the regression model.
10.6 STEPWISE REGRESSION
Stepwise regression is a variation of regular multiple regression that was invented to
speciically address the issue of variables that overlap a lot in the information they
provide about the Y. We noted above that when variables overlap a lot in what they
tell us about Y, it is possible that two, or more, variables that are not statistically sig-
niicant may nevertheless be important in the prediction process.