Dear all,
when performing Command margins and interactplot, I find the matrix result always show "not estimable". And Command interactplot doesn't run at all and remains still.
My Stata codes are shown below:
global xlist shadow betteredu engshadow schoollevel
mixed stdeng $xlist i.gender i.hukou i.single i.schoolstatus i.ecolevel i.scoreexp i.pressure i.helpstudy i.title i.city i.attitude i.ecolevel#talkstudy i.title#talkstudy c.schoollevel#talkstudy i.scoreexp#talkstudy talkstudy, ||schids: i.schoolstatus i.city schoollevel
margins i.ecolevel# talkstudy, atmeans
interactplot, term(1) name(term1, replace) scheme(tufte)
The regression results are with:
note: 4.ecolevel#1.talkstudy omitted because of collinearity.
note: 4.title#1.talkstudy omitted because of collinearity.
note: 4.scoreexp#1.talkstudy omitted because of collinearity.
Could collinearity exist as a possible reason? Or any misinterpretations of these two Commands?
Thanks to all.
Li
when performing Command margins and interactplot, I find the matrix result always show "not estimable". And Command interactplot doesn't run at all and remains still.
My Stata codes are shown below:
global xlist shadow betteredu engshadow schoollevel
mixed stdeng $xlist i.gender i.hukou i.single i.schoolstatus i.ecolevel i.scoreexp i.pressure i.helpstudy i.title i.city i.attitude i.ecolevel#talkstudy i.title#talkstudy c.schoollevel#talkstudy i.scoreexp#talkstudy talkstudy, ||schids: i.schoolstatus i.city schoollevel
margins i.ecolevel# talkstudy, atmeans
interactplot, term(1) name(term1, replace) scheme(tufte)
The regression results are with:
note: 4.ecolevel#1.talkstudy omitted because of collinearity.
note: 4.title#1.talkstudy omitted because of collinearity.
note: 4.scoreexp#1.talkstudy omitted because of collinearity.
Could collinearity exist as a possible reason? Or any misinterpretations of these two Commands?
Thanks to all.
Li
Comment