Topic Review - Newest First (only newest 5 are displayed)

docamna

Re: Re: How to calculate univariant and multivariant logistic regression anlaysis?

I would agree with Dr Rehan Qayyum!But for biggners \" SPSS\" 12 to onwards are more user friendly and one can use it.

[Edited by docamna on 27-04-2008 at 04:21 PM GMT]

ibrahim_ap

Re: How to calculate univariant and multivariant logistic regression anlaysis?

V-15 of spss is very well doing this, medcalc is also very well to do so. you can use NCCS for such also.

rqayyum

Re: Re: Re: How to calculate univariant and multivariant logistic regression anlaysis?

I agree, you are right.

siddiqi

Re: Re: How to calculate univariant and multivariant logistic regression anlaysis?

I beleive version 12 onwards SPSS can perform conditional logistic regression.

rqayyum

Re: How to calculate univariant and multivariant logistic regression anlaysis?

Well, probably the best free (and even among commercial) statistical program for both linear and logistic regression is 'R'. SPSS can do both linear and logistic regressions. It can't do conditional logistic regression. STATA is a much better program than SPSS for statistical modelling. If you are interested in doing Bayesian linear or logistc regression you can use BUGS or WinBUGS. There is a new Bayesian program called JAGS which has some additional flexibility in modelling. However, the most important thing, important than knowing which software to use is to know what is linear and logistic regression, underlying assumptions and limitations, and how to infer from the results printed out by a software program.