# st: Bivariate probit with sample selection

 From Bo Wandschneider To stata Subject st: Bivariate probit with sample selection Date Thu, 20 Feb 2003 11:02:52 -0500

We are new to Stata and are still learning our way around this package - different philosophy than things like SAS, TSP, Shazam or Limdep. We have a user who wants to duplicate some results they had in Limdep.

In limdep we simply ran the individual probits, stored the betas and used these as starting values in the bivariate probit. After this equation we entered the Select command for sample selection. We can get to the bivariate probit (without the starting values), but are unsure how to program the next step. We have checked manuals, but don't have the 'SVY' manual (we are ordering it)
Any suggestions were to look ... just need a starting point -

here is an excerpt from limdep

Probit ;Lhs = public
;Rhs = one, age, agesq, reg1, reg3, reg4, reg5, urban,
ottawa, mover, sch2, sch3, sch4, sch5, edart, human,
social, bus, hsc, sceng, mteng, mtfr, mtef, ysm, ysmsq, vismin ;
Hold results \$
list; B\$
matrix; B1=B\$
Probit ; Lhs = engfr
;Rhs = one, age, agesq, reg1, reg3, reg4, reg5, urban,
sch2, sch3, sch4, sch5, edart, human, social, bus, hsc, sceng,
mteng, mtfr, mtef, marit, que, west, us, eur, asia, ysm, ysmsq ;
Hold results \$
list; B \$
Matrix; B2=B \$

Bivariate Probit; LHS = Public, engfr;
RH1 = one, age, agesq, reg1, reg3, reg4, reg5, urban, ottawa,
mover, sch2, sch3, sch4, sch5, edart, human,
social, bus, hsc, sceng, mteng, mtfr, mtef, ysm, ysmsq, vismin ;
RH2 = one, age, agesq, reg1, reg3, reg4, reg5, urban, sch2,
sch3, sch4, sch5, edart, human, social, bus, hsc, sceng,
mteng, mtfr, mtef, marit, que, west, us, eur, asia, ysm, ysmsq ;
start = B1, B2; Hold \$

SELECT ;Lhs = LNW
;Rhs = one, exp, expsq, reg1, reg3, reg4, reg5, urban,
marri, common, divor, vismin, sch2, sch3, sch4, sch5,
edart, social, human, bus, sceng, hsc, manag, super,

SELECT ;Lhs = LNW
;Rhs = one, exp, expsq, reg1, reg3, reg4, reg5, urban,
marri,
common, divor, vismin, sch2, sch3, sch4, sch5,
edart, social, human, bus, sceng, hsc, manag, super,
science, health, socsc,

SELECT ;Lhs = LNW
;Rhs = one, exp, expsq, reg1, reg3, reg4, reg5, urban,
marri,
common, divor, vismin, sch2, sch3, sch4, sch5,
edart, social, human, bus, sceng, hsc, manag, super, science,
health, socsc,
trade, in1, in2, in3, in4, in5, in6, in8, in9, in10; selection=0,1 \$
..... and so on

--
+------------------------------+
Computing and Communications Services University of Guelph