你说的是中国农村经济上的那个论文吧,得出的结果表明,常规的probit model更为合适。
手册上面有这样一句话:At the bottom of the output is a Wald test of the exogeneity of the instrumented variables. If the test statistic is not significant, there is not sufficient information in the sample to reject the null that there is no endogeneity. Then a regular probit regression may be appropriate; the point estimates fromivprobit are consistent, though those from probit are likely to have smaller standard errors.(11版手册,728-729)
你看到的那篇论文上面的豪斯曼检验不是也是得出相同的结果吗,也就是说用IV得出的结果和普通方法得出的结果是没有系统性差异的!
那么这时倾向于采用普通方法。
个人经验,仅供参考!