[R] reporting ANOVA for nested models

[peter dalgaard]

On Sep 7, 2011, at 20:57 , zugi young wrote:

> I have the following results for an ANOVA comparing two nested models. I
> wasn't sure how I am supposed to report this result in the area of
> psychology. Specifically, am I supposed to report the DF's or just the F
> ratio? I could manually calculate the degrees of freedoms, but there must be
> a reason why R does not give this information, i.e. those are not
> conventionally used in the reporting?
> 
> Any pointers would be greatly appreciated.

What do you mean by "compute" degrees of freedom? It would seem rather clear from the output that the F test is on (1, 373) d.f. (the denominator d.f. is taken from the largest of the two models).

As for what to reporting the d.f., some journals do like to see them, mostly to allow the reader to sanity-check the results. In models which contain variance components (or should have contained them), the denominator d.f. can be revealing. As can they if you committed everyone's favorite blunder, category codes used as quantitative variables.

>> anova(fit1, fit2)
> Analysis of Variance Table
> 
> Model 1: fit1
> Model 2: fit2
>  Res.Df    RSS Df Sum of Sq      F    Pr(>F)
> 1    373 19.908
> 2    374 30.717 -1   -10.809 202.53 < 2.2e-16 ***
> ---
> Signif. codes:  0 Œ***‚ 0.001 Œ**‚ 0.01 Œ*‚ 0.05 Œ.‚ 0.1 Œ ‚ 1
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

-- 
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com
"Døden skal tape!" --- Nordahl Grieg