For the latest version of JMP Help, visit

Publication date: 11/10/2021

The Logistic Fit Report

When you fit a model using the Nominal Logistic or Ordinal Logistic personality, you obtain a Nominal Logistic Fit report or an Ordinal Logistic Fit report, respectively. By default, these reports contain the following reports:

Effect Summary

An interactive report that enables you to add or remove effects from the model. See Effect Summary Report.

Logistic Plot

(Available only if the model consists of a single continuous effect.) The logistic probability plot illustrates what the logistic model is fitting. At each value on the horizontal axis, the probability scale in the vertical direction is partitioned into probabilities for each response category. The probabilities are measured as the vertical distance between the curves, with the total across all response category probabilities summing to 1.

The points in the logistic plot represent the observations from the data table. The horizontal position of each point is determined by its value of continuous factor. The vertical position of each point is randomly chosen to be between curves that correspond to the value of its response category. Because a fixed random seed is used, the vertical positions do not differ across multiple fits of the same model.


(Available only in the Nominal Logistic personality.) After launching Fit Model, an iterative estimation process begins and is reported iteration by iteration. After the fitting process completes, you can open the Iteration History report and see the iteration steps.

Whole Model Test

Shows tests that compare the whole-model fit to the model that omits all the regressor effects except the intercept parameters. The test is analogous to the Analysis of Variance table for continuous responses. For more information about the Whole Model Test report, see Whole Model Test.

Fit Details

Shows various measures of fit for the model. See Fit Details.

Lack of Fit

(Available only when there are replicated points with respect to the X effects and the model is not saturated.) Shows a lack of fit test, also called a goodness of fit test, that addresses whether more terms are needed in the model. See Lack of Fit Test.

Parameter Estimates

Shows the parameter estimates, standard errors, and associated hypothesis tests. The Covariance of Estimates report gives the variances and covariances of the parameter estimates.

Note: The Covariance of Estimates report appears only for nominal response variables, not for ordinal response variables.

Singularity Details

(Appears only when there are linear dependencies among the model terms.) Shows a report that contains the linear functions that the model terms satisfy.

Effect Likelihood Ratio Tests

The likelihood ratio chi-square tests are calculated as twice the difference of the log-likelihoods between the full model and the model constrained by the hypothesis to be tested. The constrained model is the model that does not contain the effect. These tests can take time to do because each test requires a separate set of iterations.

Note: Likelihood ratio tests are the platform default if they are projected to take less than 20 seconds to complete. Otherwise, the default effect tests are Wald tests.

Want more information? Have questions? Get answers in the JMP User Community (