Box-Cox Results

Click the OK button in the Box-Cox Transformation dialog box to display the Box-Cox Results dialog box. See also, Box-Cox Transformation Overview and Technical Notes.

Summary. Click this button to create two summary spreadsheets for the Box-Cox transformation. The first spreadsheet contains the original and transformed variables, while the second spreadsheet displays the estimated values of transformation parameter for each and every variable, the value of the shift parameter , the mean and standard deviation of the transformed variables as well as the confidence limits for the estimated value of . The last column of the spreadsheet also includes the equation for the Box-Cox transformation. As an example, ((v1+1.0)^(3.15)-1)/(3.15), indicates a Box-Cox transformation for variable v1 with shift and transformation parameters =3.15 and =1.0 respectively.

Histograms and normal probability plots. Click this button to create histograms of the original and transformed variables as well as their normal probability plots. These graphs provide you with an effective tool for visualizing the distribution of the original and transformed variables.

Search history plots. Click this button to generate a search history plot for each and every individual variable. The search plot displays the expected normal values for each value of as tried by the search algorithm.

Number of variables to add. Use this option to adjust the number of variables to be written back to the input spreadsheet. If you set this value, for example, to 2, two new columns will be added to the input spreadsheet when you click the Add variables button. Note that the maximum number of variables to add is equal to the number of transformed variables itself.

Add variables. Click this button after setting the value in the Number of variables to add option to add new columns to the input spreadsheet.

Write back to input spreadsheet. Click this button to add the transformed variables back to the original input spreadsheet. To do so first adjust the Number of variables to add (option described above) and then click the Add variables button (also described above).