Within factor degrees of freedom for logistic regression

[Emily Janio]

Hi all,

I see the between factor degrees of freedom after running a logistic regression. Is there an option to display within factors?

[Joseph Coveney]

Do you mean ANOVA? Within- and between-subjects degrees of freedom?

The only degrees of freedom that logit would have is for the model as a whole. logit assumes asymptotic conditions, and so what would have been error degrees of freedom is take to be infinite, if that's what you're intending. If you're interested in the degrees of freedom for a main-effects (joint-test) contrast that involves a categorical explanatory variable with k categories, then it's k - 1.

.ÿ
.ÿversionÿ16.1

.ÿ
.ÿclearÿ*

.ÿ
.ÿsetÿseedÿ`=strreverse("1575283")'

.ÿ
.ÿquietlyÿsetÿobsÿ250

.ÿgenerateÿbyteÿoutÿ=ÿmod(_n,ÿ2)

.ÿ
.ÿgenerateÿbyteÿpr1ÿ=ÿruniformint(1,ÿ5)

.ÿgenerateÿdoubleÿpr2ÿ=ÿruniform(-0.5,ÿ0.5)

.ÿ
.ÿlogitÿoutÿi.pr1ÿc.pr2,ÿnolog

LogisticÿregressionÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿNumberÿofÿobsÿÿÿÿÿ=ÿÿÿÿÿÿÿÿ250
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿLRÿchi2(5)ÿÿÿÿÿÿÿÿ=ÿÿÿÿÿÿÿ3.04
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿProbÿ>ÿchi2ÿÿÿÿÿÿÿ=ÿÿÿÿÿ0.6942
Logÿlikelihoodÿ=ÿ-171.76804ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿPseudoÿR2ÿÿÿÿÿÿÿÿÿ=ÿÿÿÿÿ0.0088

------------------------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿoutÿ|ÿÿÿÿÿÿCoef.ÿÿÿStd.ÿErr.ÿÿÿÿÿÿzÿÿÿÿP>|z|ÿÿÿÿÿ[95%ÿConf.ÿInterval]
-------------+----------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿpr1ÿ|
ÿÿÿÿÿÿÿÿÿÿ2ÿÿ|ÿÿÿ.1635493ÿÿÿ.4523668ÿÿÿÿÿ0.36ÿÿÿ0.718ÿÿÿÿ-.7230734ÿÿÿÿ1.050172
ÿÿÿÿÿÿÿÿÿÿ3ÿÿ|ÿÿÿ.0073453ÿÿÿ.4638674ÿÿÿÿÿ0.02ÿÿÿ0.987ÿÿÿÿÿ-.901818ÿÿÿÿ.9165086
ÿÿÿÿÿÿÿÿÿÿ4ÿÿ|ÿÿÿÿÿ.49759ÿÿÿ.4283603ÿÿÿÿÿ1.16ÿÿÿ0.245ÿÿÿÿ-.3419809ÿÿÿÿ1.337161
ÿÿÿÿÿÿÿÿÿÿ5ÿÿ|ÿÿÿ.2985482ÿÿÿ.4651266ÿÿÿÿÿ0.64ÿÿÿ0.521ÿÿÿÿ-.6130832ÿÿÿÿÿ1.21018
ÿÿÿÿÿÿÿÿÿÿÿÿÿ|
ÿÿÿÿÿÿÿÿÿpr2ÿ|ÿÿÿ.3361663ÿÿÿ.4542857ÿÿÿÿÿ0.74ÿÿÿ0.459ÿÿÿÿ-.5542172ÿÿÿÿÿ1.22655
ÿÿÿÿÿÿÿ_consÿ|ÿÿ-.2297778ÿÿÿ.3577937ÿÿÿÿ-0.64ÿÿÿ0.521ÿÿÿÿ-.9310405ÿÿÿÿÿ.471485
------------------------------------------------------------------------------

.ÿ
.ÿtestparmÿi.pr1

ÿ(ÿ1)ÿÿ[out]2.pr1ÿ=ÿ0
ÿ(ÿ2)ÿÿ[out]3.pr1ÿ=ÿ0
ÿ(ÿ3)ÿÿ[out]4.pr1ÿ=ÿ0
ÿ(ÿ4)ÿÿ[out]5.pr1ÿ=ÿ0

ÿÿÿÿÿÿÿÿÿÿÿchi2(ÿÿ4)ÿ=ÿÿÿÿ2.39
ÿÿÿÿÿÿÿÿÿProbÿ>ÿchi2ÿ=ÿÿÿÿ0.6636

.ÿ
.ÿdisplayÿinÿsmclÿasÿtextÿ"dfÿ=ÿ"ÿr(df)
dfÿ=ÿ4

.ÿ
.ÿexit

endÿofÿdo-file


.