... (which is significant). Because these redundant variables are not statistically significant and make it difficult to estimate those that are, we remove them from the model. Remove variables one at a time to make sure that you don’t omit a useful variable that was disguised by collinearity. The table shows a summary of the simplified regression model with three explanatory variables.

The collinear explanatory variables contributed little to the explained variation: R2 falls without these (from 0.7558 to 0.7521), but the adjusted R2 grows a bit (from 0.7415 to 0.7450). The remaining estimated slopes are statistically significant and more precisely ...

Get Statistics for Business: Decision Making and Analysis, 3rd Edition now with the O’Reilly learning platform.

O’Reilly members experience live online training, plus books, videos, and digital content from nearly 200 publishers.