一般回归不用标明weight也可以吧，为什么ivtobit这么奇怪呢

*Setup
   webuse laborsup

  *  Obtain full ML estimates
    ivtobit fem_inc fem_educ kids (other_inc = male_educ), ll
    ivtobit fem_inc fem_educ kids (other_inc = male_educ), ll(12)

    *Obtain two-step estimates
      ivtobit fem_inc fem_educ kids (other_inc = male_educ), ll twostep
      ivtobit fem_inc fem_educ kids (other_inc = male_educ), ll(12) twostep

你命令写对了吗
先试试人家自带的例子

如果例子都不能运行，就是软件升级的问题

命令写的有点问题，后来试了前面的括号去掉就没事了，谢谢版主

请问那个option  twostep是什么意思？估计方法不一样吗？

看最下面，帮助的介绍
Title

    [R] ivtobit -- Tobit model with continuous endogenous regressors


Syntax

    Maximum likelihood estimator

        ivtobit depvar [varlist1] (varlist2 = varlist_iv) [if] [in] [weight], ll[(#)] ul[(#)] [mle_options]


    Two-step estimator

        ivtobit depvar [varlist1] (varlist2 = varlist_iv) [if] [in] [weight], twostep ll[(#)] ul[(#)] [tse_options]


    mle_options                 Description
    -------------------------------------------------------------------------------------------------------------------------
    Model
    * ll[(#)]                   lower limit for left censoring
    * ul[(#)]                   upper limit for right censoring
      mle                       use conditional maximum-likelihood estimator; the default
      constraints(constraints)  apply specified linear constraints

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

    Reporting
      level(#)                  set confidence level; default is level(95)
      first                     report first-stage estimates
      nocnsreport               do not display constraints
      display_options           control column formats, row spacing, line width, and display of omitted variables and base
                                  and empty cells

    Maximization
      maximize_options          control the maximization process

      coeflegend                display legend instead of statistics
    -------------------------------------------------------------------------------------------------------------------------
    * You must specify at least one of ll[(#)] and ul[(#)].

    tse_options                 Description
    -------------------------------------------------------------------------------------------------------------------------
    Model
    * twostep                   use Newey's two-step estimator; the default is mle
    * ll[(#)]                   lower limit for left censoring
    * ul[(#)]                   upper limit for right censoring

    SE/Robust
      vce(vcetype)              vcetype may be twostep, bootstrap, or jackknife

    Reporting
      level(#)                  set confidence level; default is level(95)
      first                     report first-stage estimates
      display_options           control column formats, row spacing, line width, and display of omitted variables and base
                                  and empty cells

      coeflegend                display legend instead of statistics
    -------------------------------------------------------------------------------------------------------------------------
    * twostep is required.  You must specify at least one of ll[(#)] and ul[(#)].

    varlist1 and varlist_iv may contain factor variables; see fvvarlist.
    depvar, varlist1, varlist2, and varlist_iv may contain time-series operators; see tsvarlist.
    bootstrap, by, jackknife, rolling, statsby, and svy are allowed; see prefix.
    Weights are not allowed with the bootstrap prefix.
    vce(), first, twostep, and weights are not allowed with the svy prefix.
    fweights, iweights, and pweights are allowed with the maximum likelihood estimator.  fweights are allowed with Newey's
      two-step estimator.  See weight.
    coeflegend does not appear in the dialog box.
    See [R] ivtobit postestimation for features available after estimation.


Menu

    Statistics > Endogenous covariates > Tobit model with endogenous covariates


Description

    ivtobit fits tobit models where one or more of the regressors is endogenously determined.  By default, ivtobit uses
    maximum likelihood estimation. Alternatively, Newey's (1987) minimum chi-squared estimator can be invoked with the
    twostep option. Both estimators assume that the endogenous regressors are continuous and so are not appropriate for use
    with discrete endogenous regressors.  See [R] ivprobit for probit estimation with endogenous regressors and [R] tobit for
    tobit estimation when the model contains no endogenous regressors.