|
|
Dear SPSS listserv members,
My name is Fernando Biscione, MD, postgraduate student from Brazil. I'm interested in using binary logistic regression analysis for my thesis. My purpose is to use logistic regression in a cohort of surgical patients to describe independent risk factors for infection. I'd like very much to ask you some questions:
1) if my primary interest is to independent describe risk factors, and not to construct a model to predict a dependent variable on the basis of independents, does it really matter for my analysis the Hosmer and Lemeshow goodness-of-fit and the analysis of residuals?;
2) most (except for one) independents in my analysis are binary, is multicollinearity a potential problem in this case?, and additionaly, can I assume linearity between the independents and the logit of the dependent without further analysis?.
3) if the standar errors of the b coefficients are all small, does it straighforward exclude multicollinearity among the independents without the need for further analysis?.
Thanks in advance,
King regards,
Fernando Biscione, MD.
---------------------------------
1GB gratis, Antivirus y Antispam
Correo Yahoo!, el mejor correo web del mundo
Abrí tu cuenta aquí
---------------------------------
Horóscopos, Salud y belleza, Chistes, Consejos de amor.
El contenido más divertido para tu celular está en
Yahoo! Móvil
|
Locked
