PRB Instruction

PRBIT(options) depvar start end probs coeffs

# list of explanatory variables in Regression Format

Estimates a probit model. The newer DDV instruction can handle logit, probit, and other variations on these models. We generally recommend using DDV rather than the older PRB.

Parameters

depvar	dependent variable with values coded ad zeros and non-zeros
start, end	range for estimation, defaults to maximum range possible given variables
probs	series for probabilities of achieving observed values

Options

Standard Regression Options

Standard Non-Linear Estimation Options

Robust Error Options

EQUATION=Equation to estimate [unused]

INITIAL=VECTOR of initial estimates [zeros]

Variables Defined

%BETA	coefficient vector (VECTOR)
%XX	Estimated covariance matrix (SYMMETRIC)
%STDERRS	vector of coefficient standard errors (VECTOR)
%TSTATS	vector of t-statistics of the coefficients (VECTOR)
%NOBS	number of observations (INTEGER)
%NREG	number of regressors (INTEGER)
%LOGL	log likelihood value (REAL)
%NCORRECT	number of cases correct (INTEGER)
%FUNCVAL	final value of the function being maximized. (REAL)
%ITERS	iterations completed (INTEGER)
%CONVERGED	=1 or 0. 1 indicates that the process converged.
%CONVCRIT	final convergence criterion (REAL). This will be equal to zero if the sub-iteration limit was reached on the last iteration.
%RSQUARED	pseudo \(R^2\) measure (REAL)

Example

prb yesvote

# constant educat income state age

The probit model is the default model for DDV, so this could also be done by substituting DDV for PRB as the command.