logit model without constant: ''no convergence''?

#43
Dear Greta. You pushed me to try regression, thanks for it. Reorganizing the data by original observations was extra good resulted, thanks to you. Im satisfied both regression and logit models. Thanks for all helps..
Odds ratios and the rest are all the same in interpretting, said hlsmith.
For now, do you recommend me comparing both models -regression and firthlogit- in my study? Or should i use only one..
 
#44
Could you also please lead about correlating original c/b ratios (cont,nuous) and categorical explanatories.
Thanks for everything till now..
 

hlsmith

Omega Contributor
#45
Funny enough. I did not know you could input data like Greta showed in post #7. That entry method is going to save me a lot of hassle. I had previously only recallwed the method where you use many "' around about everything.


:)
 
#46
i wasnt let to show the data, sir. So i couldnt ever think regression.
Thanks for all helps. These were the best ideas i've ever heard and i'll use all of these in my study.
Thanks again.