[R] dropterm() in MASS

[John Fox]

Dear Patrick,

At 11:51 AM 4/29/2002 +1200, Patrick Connolly wrote:

>To compare two different models, I've compared the result of using
>dropterm() on both.
>
>
>Single term deletions
>
>Model:
>growth ~ days + I(days^0.5)
>             Df Sum of Sq     RSS     AIC
><none>                    2.8750 -0.2290
>days         1    4.8594  7.7344  4.6984
>I(days^0.5)  1    0.0234  2.8984 -2.1722
>
>AND
>
>Single term deletions
>
>Model:
>growth ~ days + I(days^2)
>           Df Sum of Sq     RSS     AIC
><none>                  2.8750 -0.2290
>days       1    5.8338  8.7088  5.5290
>I(days^2)  1    0.0234  2.8984 -2.1722
>
>I understood that the AIC column was the change in AIC when that term
>was dropped from the model.
>
>QUESTION:
>
>Why would (should) the effect of dropping quite different terms be
>exactly the same?  It's consistent with all the data of this type I've
>tried it on.

Notice that the residual sums of squares for the two models are identical, 
so the peculiar result here pertains not just to dropping a term but also 
to the general fit of the model. What does the variable days look like? Are 
there just three different values of this variable?

I hope that this helps,
  John



-----------------------------------------------------
John Fox
Department of Sociology
McMaster University
Hamilton, Ontario, Canada L8S 4M4
email: jfox at mcmaster.ca
phone: 905-525-9140x23604
web: www.socsci.mcmaster.ca/jfox
-----------------------------------------------------

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._