Example of the Fit Robust Option
This example shows you how use the Bivariate platform to fit a robust model using the Huber M-estimation method.
1. Select Help > Sample Data Folder and open Weight Measurements.jmp.
2. Select Analyze > Fit Y by X.
3. Select weight and click Y, Response.
4. Select height and click X, Factor.
5. Click OK.
6. Click the Bivariate Fit red triangle and select Fit Line.
7. Click the Bivariate Fit red triangle and select Robust > Fit Robust.
Figure 5.18 Example of Robust Fit
Three measurements are marked by stars in the scatterplot. These three observations appear to have weights that are lower than expected. When fitting a regression line (denoted by the red line), these points have leverage on the fit. The p-value of 0.1203 in the Analysis of Variance report suggests a weak linear relationship between height and weight. However, when you consider the Robust Fit (denoted by the green line), the linear relationship between height and weight is stronger than for the standard fit. The Robust Fit p-value of 0.0488 supports the hypothesis of a linear relationship between height and weight. The measurements identified with unusually low weights were influencing the analysis.