Regression Estimation Options

The regression method estimates missing values using multiple linear regression. The means, the covariance matrix, and the correlation matrix of the predicted variables are displayed.

Estimation Adjustment. The regression method can add a random component to regression estimates. You can select residuals, normal variates, Student's t variates, or no adjustment.

  • Residuals. Error terms are chosen randomly from the observed residuals of complete cases to be added to the regression estimates.
  • Normal Variates. Error terms are randomly drawn from a distribution with the expected value 0 and the standard deviation equal to the square root of the mean squared error term of the regression.
  • Student's t Variates. Error terms are randomly drawn from a t distribution with the specified degrees of freedom, and scaled by the root mean squared error (RMSE).

Maximum number of predictors. Sets a maximum limit on the number of predictor (independent) variables used in the estimation process.

Save completed data. Writes a dataset in the current session or an external IBM® SPSS® Statistics data file, with missing values replaced by values estimated by the regression method.

To Specify Regression Options

This feature requires the Missing Values option.

  1. From the menus choose:

    Analyze > Missing Value Analysis...

  2. In the main Missing Value Analysis dialog box, select the variable(s) for which you want to estimate missing values using the regression method.
  3. Select Regression in the Estimation group.
  4. To specify predicted and predictor variables, click Variables. See the topic Predicted and Predictor Variables for more information.
  5. Click Regression.
  6. Select the regression options you want.