Model Fitting: Linear Regression

Transforming the Response

The salary variable ranges from 67.5 to 2460 (measured in thousands of dollars). Since the variation of salaries is much greater for the higher salaries, it is appropriate to apply a logarithmic transformation to the salaries before fitting the model. You can use the Variable Transformation Wizard to transform the salary variable, as described in Chapter 32, "Variable Transformations."


Select Analysis \blacktriangleright\,Variable Transformation from the main menu.

The Variable Transformation Wizard in Figure 21.1 appears.


Select the log10(Y+a) transformation from the Transformations list.



ugfitregtransform1.png (8840 bytes)

Figure 21.1: Selecting a Log10 Transformation


Click Next.

The wizard displays the page shown in Figure 21.2.

ugfitregtransform2.png (10955 bytes)

Figure 21.2: Selecting a Variable and Parameters


Scroll to the end of the variable list. Select the salary variable, and click Set Y.

The parameter a is an offset that is useful if your variable contains nonpositive values. For these data, you can accept the default value of .


Click Finish.

Because there are missing values for the salary variable, a warning message appears (Figure 21.3) informing you that the transformed values for these observations are set to missing values.

ugfitregtransformmsg.png (4496 bytes)

Figure 21.3: A Warning Message


Click OK to dismiss the warning message.

Stat Studio adds the new variable, Log10_salary, as the last variable in the data set.

Previous Page | Next Page | Top of Page