Stepwise Regression with JMP
What is Stepwise Regression?
Stepwise regression is a statistical method to automatically select regression models with the best sets of predictive variables from a large set of potential variables. There are different statistical methods used in stepwise regression to evaluate the potential variables in the model:
- F-test
- T-test
- R-square
- AIC
Three Approaches to Stepwise Regression
- Forward Selection
Bring in potential predictors one by one and keep them if they have a significant impact on improving the model. - Backward Selection
Try out potential predictors individually and eliminate them if they are insignificant to improve the fit. - Mixed Selection
Is a combination of both forward selection and backward selection. Add and remove variables based on pre-defined significance threshold levels.
How to Use JMP to Run a Stepwise Regression
Case study: We want to build a regression model to predict the oxygen uptake of a person who runs 1.5 miles. The potential predictors are:
- Age
- Weight
- Runtime
- Runpulse
- RstPulse
- MaxPulse
Data File: “Stepwise Regression.jmp”
Run Stepwise Regression in JMP:
- Click Analyze -> Fit Model
- Model Specification window appears.
- Select “Oxy” as the Y and add the potential factors to the model effects box
Select “Stepwise” in the “Personality” dropdown box - Click “Run Mode
- The “Stepwise Fit” page shows up
- Select the P-value Threshold for Stopping Rule
- Enter the “Prob to Enter” and “Prob to leave” thresholds into the corresponding text boxes.
- Select the stepwise regression direction
- Forward
- Backward
- Mixed
- Click the “Go” button to let JMP automatically find the set of predictors satisfying the pre-defined significance probability thresholds.
Model summary: Two of the seven potential factors are not statistically significant since their p-value is higher than the alpha to enter. Step History: Step-by-step records on how to develop the final model. Each column indicates the model built in each step.
