Lack of Fit Test

The Lack of Fit test addresses whether there is enough information in the current model or whether more complex terms are needed. This test is sometimes called a goodness-of-fit test. The lack of fit test calculates a pure-error negative log-likelihood by constructing categories for every combination of the model effect values in the data. The Saturated row in the Lack Of Fit table contains this log-likelihood. The Lack of Fit report also contains a test of whether the Saturated log-likelihood is significantly better than the Fitted model.

The Saturated degrees of freedom is m–1, where m is the number of unique populations. The Fitted degrees of freedom is the number of parameters not including the intercept.

The Lack of Fit table contains the negative log-likelihood for error due to Lack of Fit, error in a Saturated model (pure error), and the total error in the Fitted model. The chi-square statistic tests for lack of fit.