用stata进行因变量为二元分布的多层次非线性模型(多层次logistic)分析的时候:stata默认只给出sd(_cons),而不会像多层次线性模型同时也给出sd(Residual);那么在多层次非线性模型分析中怎么计算组内相关系数呢?如:运行 xtmelogit poorhealth ||communityID: ,variance or
得到:
Mixed-effects logistic regression Number of obs = 14019
Group variable: communityID Number of groups = 434
Obs per group: min = 10
avg = 32.3
max = 83
Integration points = 7 Wald chi2(0) = .
Log likelihood = -9591.5311 Prob > chi2 = .
------------------------------------------------------------------------------
poorhealth | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
_cons | 1.092675 .029656 3.27 0.001 1.036069 1.152373
------------------------------------------------------------------------------
------------------------------------------------------------------------------
Random-effects Parameters | Estimate Std. Err. [95% Conf. Interval]
-----------------------------+------------------------------------------------
communityID: Identity |
var(_cons) | .1788416 .0223419 .1400011 .2284575
------------------------------------------------------------------------------
LR test vs. logistic regression: chibar2(01) = 214.30 Prob>=chibar2 = 0.0000
多层次logistic分析中,组内相关系数依然是等于var(_cons)/[var(_cons)+var(Residual)] 吧?