------------------------------------------------------------------------------- help formvppredCappellari and Jenkins (31oct2002) -------------------------------------------------------------------------------

Predictions from multivariate probit models estimated by SML

mvpprednewvarname_prefix[ifexp] [inrange] [,statistic]where

statisticis one of

xbthe linear prediction for each equation; the default.stdpthe standard error of the linear predictions for each equation.pmargthe marginal success probability for each equation.pallthe joint probabilities: (i) Pr(depvar_i= 1, alli) and (ii) Pr(depvar_i= 0, alli), for equationsi= 1,...,M.These statistics are available both in and out of sample; type "

predict...ife(sample)..." if wanted only for the estimation sample. If no statistic is specified, the default isxb.

Description

mvppredprovides predictions following estimation of anM-equation probit model by the method of simulated maximum likelihood (SML) using the programmvp7. See mvp7. Predictions provided for each equation are the fitted index values, the standard errors of the fitted index values for each equation, the predicted marginal success probabilities, and two predicted joint probabilities. These are: (i) Pr(depvar_i= 1, alli) and (ii) Pr(depvar_i= 0, alli), for equationsi= 1,...,M. (Additional joint probabilities and conditional probabilities have not been provided because the number of such probabilities increases substantially asMincreases.) The multivariate normal distributions used to calculate the joint probabilities are derived by simulation using the GHK simulator, with the seed and number of random draws the same as used bymvp7to derive the parameter estimates.

Options

xb, the default, calculates the linear prediction (Xb) for each equation. Results are stored in the variablesnewvarname_prefixi, for equationsi= 1,...,M.

stdpcalculates the standard error of the linear prediction (Xb) for each equation. Results are stored in the variablesnewvarname_prefixi, for equationsi= 1,...,M.

pmargcalculates the marginal probit predicted probability of success for each outcome, Pr(depvari) = 1, for each equationi= 1,...,M. Results are stored in the variablesnewvarname_prefixi, for equationsi= 1,...,M.

pallcalculates (i) the probit predicted joint probability of success in every outcome, Pr(depvari) = 1, for alli= 1,...,M, and (ii) the probit joint probability of failure in every outcome, Pr(depvari) = 0, for alli= 1,...,M. Results are stored in the variablesnewvarname_prefix1sfor predicted probability (i) andnewvarname_prefix0sfor predicted probability (ii).

Examples. use http://www.stata-press.com/data/r7/school.dta, clear

. mvp7 (private = years logptax loginc) (vote = years logptax loginc)

. mvppred xb

. mvppred pall, pall

. mvppred pmarg, pmarg

. mvppred stdp, stdp

. sum xb1, xb2, pmarg1, pmarg2, stdp1, stdp2, pall1s pall0s

AuthorsLorenzo Cappellari, Universita del Piemonte-Orientale, Italy <Lorenzo.Cappellari@eco.unipmn.it>

Stephen P. Jenkins, ISER, University of Essex, U.K. <stephenj@essex.ac.uk>

Also seeManual:

[U] 23 Estimation and post-estimation commands,[U] 29 Overview of model estimation in Stata,On-line: help for mvp7, postest, and predict.