紧急求助，关于BIPROBIT

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

小弟正在做一个BIVARIATE PROBIT的分析，但是有一个问题搞不清楚，紧急求助各位大牛
在BIVARIATE PROBIT中， 需要加入IDENTIFYING RESTRICTION变量，这个变量是只能在SELECTION EQUATION中加呢，还是
也可以通过在OUTCOME EQUATION中加入来解决？

非常感谢！！

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏3 回帖

关键词：biprobit Probit 紧急求助 Rob IPR 求助 biprobit

你查过biprobit的帮助或者手册吗
好像没有你说的那种

下面是stata里面的help文件


help biprobit                                              dialogs:  biprobit                        
                                                                     seemingly unrelated biprobit     
                                                                     svy: biprobit                    
                                                                     svy: seemingly unrelated biprobit
                                                          also see:  biprobit postestimation         
------------------------------------------------------------------------------------------------------

Title

    [R] biprobit -- Bivariate probit regression


Syntax

    Bivariate probit model

        biprobit depvar1 depvar2 [varlist] [if] [in] [weight] [, options]


    Seemingly unrelated bivariate probit model

        biprobit equation1 equation2 [if] [in] [weight] [, su_options]


    where equation1 and equation2 are specified as

        ( [eqname: ] depvar [=] [varlist] [, noconstant offset(varname) ] )

    options                       description
    ------------------------------------------------------------------------------------------------
    Model
      noconstant                  suppress constant term
      partial                     fit partial observability model
      offset1(varname)            offset variable for first equation
      offset2(varname)            offset variable for second equation
      constraints(constraints)    apply specified linear constraints
      collinear                   keep collinear variables

    SE/Robust
      vce(vcetype)                vcetype may be oim, robust, cluster clustvar, opg, bootstrap, or
                                    jackknife

    Reporting
      level(#)                    set confidence level; default is level(95)
      noskip                      perform likelihood-ratio test

    Max options
      maximize_options            control the maximization process; seldom used
    ------------------------------------------------------------------------------------------------

    su_options                    description
    ------------------------------------------------------------------------------------------------
    Model
      partial                     fit partial observability model
      constraints(constraints)    apply specified linear constraints
      collinear                   keep collinear variables

    SE/Robust
      vce(vcetype)                vcetype may be oim, robust, cluster clustvar, opg, bootstrap, or
                                    jackknife

    Reporting
      level(#)                    set confidence level; default is level(95)
      noskip                      perform likelihood-ratio test

    Max options
      maximize_options            control the maximization process; seldom used
    ------------------------------------------------------------------------------------------------

    depvar1, depvar2, varlist, and depvar may contain time-series operators; see tsvarlist.
    bootstrap, by, jackknife, rolling, statsby, svy, and xi are allowed; see prefix.
    Weights are not allowed with the bootstrap prefix.
    vce(), noskip, and weights are not allowed with the svy prefix.
    pweights, fweights, and iweights are allowed; see weight.
    See [R] biprobit postestimation for features available after estimation.


Description

    biprobit fits maximum-likelihood two-equation probit models -- either a bivariate probit or a
    seemingly unrelated probit (limited to two equations).


Options

        +-------+
    ----+ Model +-----------------------------------------------------------------------------------

    noconstant; see [R] estimation options.

    partial specifies that the partial observability model be fitted.  This particular model
        commonly has poor convergence properties, so we recommend that you use the difficult option
        if you want to fit the Poirier partial observability model; see [R] ml.

        This model computes the product of the two dependent variables so that you do not have to
        replace each with the product.

    offset1(varname), offset2(varname), constraints(constraints), collinear; see [R] estimation
        options.

        +-----------+
    ----+ SE/Robust +-------------------------------------------------------------------------------

    vce(vcetype) specifies the type of standard error reported, which includes types that are
        derived from asymptotic theory, that are robust to some kinds of misspecification, that
        allow for intragroup correlation, and that use bootstrap or jackknife methods; see [R]
        vce_option.

        +-----------+
    ----+ Reporting +-------------------------------------------------------------------------------

    level(#); see [R] estimation options.

    noskip specifies that a full maximum-likelihood model with only a constant for the regression
        equation be fitted.  This model is not displayed but is used as the base model to compute a
        likelihood-ratio test for the model test statistic displayed in the estimation header.  By
        default, the overall model test statistic is an asymptotically equivalent Wald test of all
        the parameters in the regression equation being zero (except the constant).  For many
        models, this option can substantially increase estimation time.

        +-------------+
    ----+ Max options +-----------------------------------------------------------------------------

    maximize_options: difficult, technique(algorithm_spec), iterate(#), [no]log, trace, gradient,
        showstep, hessian, shownrtolerance, tolerance(#), ltolerance(#), gtolerance(#),
        nrtolerance(#), nonrtolerance, from(init_specs); see [R] maximize.  These options are seldom
        used.

        Setting the optimization type to technique(bhhh) resets the default vcetype to vce(opg).


Examples

    Setup
        . webuse school

    Bivariate probit model
        . biprobit private vote logptax loginc years

    Seemingly unrelated bivariate probit model
        . biprobit (private = logptax loginc years) (vote = logptax years)

    Seemingly unrelated bivariate probit model with robust standard errors
        . biprobit (private = logptax loginc years) (vote = logptax years), vce(robust)

I have used biprobit for a lot of my papers. One thing you need to know is that there is NO strict identification for biprobit. That is, you can have z from the selection equation and x from the outcome equation to be the same: x = z. The system is still identified because it is nonlinear. So the identification comes from the nonlinear property of bivariate probit. If this is the case, then the underlying assumption is that the error terms in the two equations are correlated.

However, in practice, people still prefer to have one or two independent variables in the selection equations to be different from the outcome equation. Z not = X. This is sometimes called strictly identified, which is a prefered approach.

You may also check out mvprobit. It is a user written ado file. It does multivariate probit: bivariate probit, trivariate probit, etc etc.

3# gyw23

Thanks, that's really a clear explanation. And, another question, is there any restriction or preference (dummy or ordinal, etc.) in terms of choosing the independent identifying variable in the selection equation? Because in my equation, i would include several dummy variables as the identifying variable and i dont know if it is a reasonable choice. Looking forward to your comments!

受教了，很好的一个命令

麻烦问问 在做biprobit模型时，估计系数的命令咋写啊？