[R] ob.step$anova interpretation

[Prof Brian Ripley]

stepAIC is support software for a book.  Please do as the posting guide 
asks and consult the book, which explains both AIC and how hte function 
works.

The book is even a reference on the help page.

On Mon, 8 May 2006, orkun wrote:

> hello
>
> I built logistic regression model.
> To model check I used stepAIC. But I don't know how it
> is interpreted . I could not any find any explanation about it
> For instance which model is preferable ? What are the critarias
> to choose beter model
>
> I will appreciate if you give me an explanation ?
>
> models
> ---------
>
> > lo1.step$anova
> Stepwise Model Path
> Analysis of Deviance Table
>
> Initial Model:
> res ~ as.factor(jeo) + as.factor(eg) + as.factor(ys) + as.factor(yy) +
>    as.factor(ak)
>
> Final Model:
> res ~ as.factor(jeo) + as.factor(eg) + as.factor(ys) + as.factor(yy) +
>    as.factor(ak)
>
>
>  Step Df Deviance Resid. Df Resid. Dev      AIC
> 1                        788   30604.46 31588.77
> ~~~~~~~~~~
> > lo2.step$anova
> Stepwise Model Path
> Analysis of Deviance Table
>
> Initial Model:
> res ~ as.factor(jeo):as.factor(eg) + as.factor(ys) + as.factor(yy) +
>    as.factor(ak)
>
> Final Model:
> res ~ as.factor(ys) + as.factor(yy) + as.factor(ak)
>
>
>                            Step Df Deviance Resid. Df Resid. Dev        AIC
> 1                                                  728 3123352.66 3124456.98
> 2 - as.factor(jeo):as.factor(eg) 76  3052176       804   71176.54   72128.85
>
>
> kind regards
>
>

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595