※ 引述《pths40 (不信蒼生)》之銘言： : 小弟才疏學淺 在此請教各位前輩 : model: : y1=alpha1+alpha2*y2+alpha3*y3+alpha4*x1 : y2=beta1+beta2*x2 : y3=gamma1+gamma2*x3 : which one could be solve by 2SLS and how? : are the estimates unbias? consistent? : 小弟想法是y1 : 1: OLS predict haty2 & haty3 : 2: OLS reg y1 haty2 haty3 x1 : unbias and consistent since it's OLS 2SLS will be consistent but not unbiased the reason is haty2 and haty3 are estimated, so it's impossible to be unbiased -- 一蕭一劍平生意 負盡狂名十八年 --

※ 發信站: 批踢踢實業坊(ptt.cc) ◆ From: 35.13.62.44