Goodness of fit tests – P-value for Probit Analysis

Use the Pearson and deviance goodness-of-fit measures to evaluate how well the selected distribution fits the data.

After running the analysis, look at the p-value for the goodness-of-fit tests.

  • Higher p-values indicate that the model fits the data well.
  • Lower p-values indicate that the predicted probabilities from the model differ significantly from the observed probabilities in the data. Thus, the model does not fit the data well. Selecting another distribution may improve the model's fit.

Unless one model has a special meaning in your discipline, you may want to run a probit analysis again using other models and select one that produces the largest goodness-of-fit p-values.

Example output

Goodness-of-Fit Tests Method Chi-Square DF P Pearson 1.19972 6 0.977 Deviance 1.22858 6 0.975 Tolerance Distribution

Interpretation

The high p-values for the windshield data (0.977 and 0.975) indicate that the chosen distribution adequately fits the data.

By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy