Lack of Fit Tests

The PROBIT Procedure

Lack of Fit Tests

Two goodness-of-fit tests can be requested from the PROBIT procedure -a Pearson chi-square test and a log-likelihood ratio chi-square test.

If there is only a single continuous independent variable, the data are internally sorted to group response values by the independent variable. Otherwise, the data are aggregated into groupings that are delimited whenever a change is observed in one of the independent variables.

Note: Because of this grouping, the data set should be sorted by the independent variables before the PROBIT procedure is run if the LACKFIT option is specified.

If the Pearson goodness-of-fit chi-square test is requested and the p-value for the test is too small, variances and covariances are adjusted by a heterogeneity factor (the goodness-of-fit chi-square divided by its degrees of freedom) and a critical value from the t distribution is used to compute the fiducial limits. The Pearson chi-square test statistic is computed as

$\sum_i \sum_j \frac{(r_{ij} - n_i p_{ij})^2}{n_i p_{ij}}$

where the sum on i is over grouping, the sum on j is over levels of response, the r_ij is the frequency of response level j for the ith grouping, n_i is the total frequency for the ith grouping, and p_ij is the fitted probability for the jth level at the ith grouping.

The log-likelihood ratio chi-square test statistic is computed as

$2 \sum_i \sum_j r_{ij} \ln ( \frac{r_{ij}}{n_i p_{ij}} )$

This quantity is sometimes called the deviance. If the modeled probabilities fit the data, these statistics should be approximately distributed as chi-square with degrees of freedom equal to (k - 1) ×m - q, where k is the number of levels of the multinomial or binomial response, m is the number of sets of independent variable values (covariate patterns), and q is the number of parameters fit in the model.

In order for the Pearson statistic and the deviance to be distributed as chi-square, there must be sufficient replication within the groupings. When this is not true, the data are sparse, and the p-values for these statistics are not valid and should be ignored. Similarly, these statistics, divided by their degrees of freedom, cannot serve as indicators of overdispersion. A large difference between the Pearson statistic and the deviance provides some evidence that the data are too sparse to use either statistic.

Chapter Contents
Previous
Next
Top