Home  /  Products  /  Stata 9  /  Multinomial probit

This page contains only historical information and is not about the current release of Stata. Please see our Stata 18 page for information on the current version of Stata.


Multinomial probit

  • New estimation command asmprobit fits multinomial probit (MNP) models to categorical data and is frequently used in choice-based modeling. asmprobit allows several correlation structures for the alternatives, including completely unstructured, where all possible correlations are estimated. It also allows for either heteroskedastic or homoskedastic variances among the alternatives and allows arbitrary patterns within the alternative variances or correlations. asmprobit’s syntax makes specifying both case-specific and alternative-in-case-specific regressors easy.

    In addition to common postestimation commands, such as mfx for computing marginal effects, new command estat provides additional statistics and results:

    • estat alternatives reports summary statistics about each of the the alternatives and provides a mapping between the index numbers labeling the alternatives and their associated values and labels in the dataset.

    • estat covariance computes and reports the estimated covariance matrix for the alternatives.

    • estat correlation reports the correlations among the alternatives in matrix form.

    Predicted statistics after asmprobit include the linear predictor, the probability an alternative is selected, and the standard error of the linear predictor.

    See [R] asmprobit and [R] asmprobit postestimation

  • New estimation command mprobit also fits multinomial probit models to categorical data but in the simplified situation of having only case-specific covariates (as with the multinomial logistic regression, mlogit). Maximizing the likelihood is much faster in such cases because the numerical approximation to the likelihood is simpler. See [R] mprobit