bfzldh 发表于 2020-5-13 01:00
3分类变量设置成2个虚拟变量就可以。2分类变量建议编码成0和1。
谢谢您的解答。
level1中既有分类变量(uncentered),也有连续变量(group centered),不知这样子进入模型是对的吗?分类变量需要centered吗?
level1中我的分类变量uncentered,连续变量group centered,在全模型中,二层变量是连续变量(grand centered),除了截距,其他都用固定效应,还是出不来结果,请问怎么解决呢?(在普通ols回归中,我检验了vif值,level1中所有自变量在之前的检验中vif值都在5以下,应该不存在严重多重共线性)
Level-1 Model
Y = B0 + B1*(AGEOPEN) + B2*(DUM31) + B3*(DUM32) + B4*(FUNDS) + B5*(TYPE2) + B6*(NASSESTS) + B7*(LEVEL5) + R
Level-2 Model
B0 = G00 + G01*(POPU_DEN) + G02*(WAGE_WAN) + G03*(GDP_WAN) + U0
B1 = G10 + G11*(POPU_DEN) + G12*(WAGE_WAN) + G13*(GDP_WAN)
B2 = G20 + G21*(POPU_DEN) + G22*(WAGE_WAN) + G23*(GDP_WAN)
B3 = G30 + G31*(POPU_DEN) + G32*(WAGE_WAN) + G33*(GDP_WAN)
B4 = G40 + G41*(POPU_DEN) + G42*(WAGE_WAN) + G43*(GDP_WAN)
B5 = G50 + G51*(POPU_DEN) + G52*(WAGE_WAN) + G53*(GDP_WAN)
B6 = G60 + G61*(POPU_DEN) + G62*(WAGE_WAN) + G63*(GDP_WAN)
B7 = G70 + G71*(POPU_DEN) + G72*(WAGE_WAN) + G73*(GDP_WAN)
Run-time deletion has reduced the number of level-1 records to 136
Run-time deletion has reduced the number of level-2 groups to 23
There is a problem in the fixed portion of the model. A near singularity is
likely. Possible sources are a collinearity or multicollinearity among the
predictors. We suggest that you examine a correlation matrix among the fixed
effect predictors.