我正在纸上进行复制研究。本文使用Tobit回归模型,它给出了模型的重要变量和Efron的$ R ^ 2 $。为了完全复制,我获得了数据集并为Stata中的Tobit回归做了建模。但是,当我使用spost9
包和fitstat
作为模型时,它只输出McFadden的$ R ^ 2 $,McFadden的Adj。 $ R ^ 2 $,ML(Cox-Snell)$ R ^ 2 $,Cragg-Uhler(Nagelkerke)$ R ^ 2 $,以及McKelvey& Zavoina的$ R ^ 2 $。
有没有办法让Stf的Tobit模型获得Efron的$ R ^ 2 $?
以下是模型的结果:
Tobit regression Number of obs = 19318
LR chi2(37) = 5524.53
Prob > chi2 = 0.0000
Log likelihood = -53623.815 Pseudo R2 = 0.0490
Measures of Fit for tobit of helpfulness
Log-Lik Intercept Only: -56386.081 Log-Lik Full Model: -53623.815
D(19279): 107247.629 LR(37): 5524.533
Prob > LR: 0.000
McFadden's R2: 0.049 McFadden's Adj R2: 0.048
ML (Cox-Snell) R2: 0.249 Cragg-Uhler(Nagelkerke) R2: 0.249
McKelvey & Zavoina's R2: 0.286
Variance of y*: 7152.903 Variance of error: 5109.649
AIC: 5.556 AIC*n: 107325.629
BIC: -83012.823 BIC': -5159.387
BIC used by Stata: 107632.512 AIC used by Stata: 107325.629