The mean of a binary variable is totally interpretable;
in this case the interpretation is not that an individual
has half a degree, but that on average half the individuals
have degrees.
Logit, probit and much else would not make sense otherwise.
Nick
n.j.cox@durham.ac.uk
Jingky P. Lozano
> I was thinking that you can probably use Kruskall-Wallis
> which is the nonpar
> equivalent of ANOVA but since you only have two possible
> answer (with Univ and
> without univ degree) per religion, I agree that you can
> just simply use chi-
> square. There will really be not much use in comparing the
> means. A 0.5 mean
> in education variable cannot be interpreted as incomplete
> Univ degree since
> your coding system is categorical. I think it is more
> appropriate to use chi-
> square.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/