prolems in a excerise, can somebody help?

[Ahmed Aziz]

Can somebody help us with that.
Propose an appropiate regression model to analyse the dependence of the incidens of new fractures on the medication and the status of pre-existing fractures. Use a model without interaction.

Our guess is "logistic regression model".

With the code:
logit fracture i.medication i.preexist [fweight=freq]

It just doesn't give the right answer :-(

[Nick Cox]

It's not explicit in the FAQ (it was, but we cut it out because we only rarely get such requests), but it's a longstanding tradition that the scope of Statalist doesn't include getting others to do your homework. If you need help with your exercises, your own institution will have support whenever that is part of the teaching program. If it's the expectation of an institution that you work on your own, it is really not in your best interests to seek help from others.

[Ahmed Aziz]

Dear Nick Cox
I fully understand the policy. The aim wasn't to let other people do the homework, but to point out, what I am getting wrong, so I can correct it and understand it.

Best regards

[Carlo Lazzaro]

Ahmed:
I do share Nick's reply.
However, if learning what went wrong with your code is your intention, you should better post your query on the General forum.
This would imply that you provide the list with other details about your research and post what you typed (as you already did) and what Stata gave you back (as per FAQ).
As an aside, reporting that a given statistical procedure did not give you back the right answer is not, in my opinion, the best way to capture others' attention.