From the course: 11 Useful Tips for Regression Analysis
Unlock the full course today
Join today to access over 22,600 courses taught by industry experts or purchase this course individually.
Centering
- [Instructor] It is common to transform variables in a regression. We can multiply or divide variables with each other to create new variables that lead to model improvement. However, we can also modify variables by simple additive transformations. This involves either adding or subtracting a constant to a variable. A common method is mean centering, where the mean of a variable is subtracted from itself. And this results in a variable that is centered on zero but will have a negative and positive values, indicating deviations away from the mean. However, centering can involve any additive transformation. For example, we can add the value 10 to every observation in the variable. And that is also a kind of centering. Why would we choose to center a variable in a regression? There are two reasons that are often highlighted. The first, is to improve the interpretation of the interceptor. The interceptor omits the value of…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
Weighted regression5m 36s
-
(Locked)
Factor variables5m 40s
-
(Locked)
Polynomial variables4m 36s
-
(Locked)
Fractional variables5m 38s
-
(Locked)
Model proportions5m 39s
-
(Locked)
Centering5m 1s
-
(Locked)
Missing data5m 55s
-
(Locked)
Standardized estimates3m 22s
-
(Locked)
Graph estimates4m 9s
-
(Locked)
Contour plots3m 37s
-
(Locked)
Animate results4m 18s
-