[S] ANOVA for GAM

Ottar N. Bjornstad (bjornsta@nceas.ucsb.edu)
Tue, 31 Mar 1998 11:55:01 +0100


Dear all,

I am confused about the following:
When I make the following model:
tmp1 <- gam(en ~ s(null, df = 2), family = quasi(link = identity, variance="mu"))
anova(tmp1, test="F")

I get the following output:

DF for Terms and F-values for Nonparametric Effects

Df Npar Df Npar F Pr(F)
(Intercept) 1
s(null, df = 2) 1 1 8.617401 0.004920457

which is a test for the significance of the nonlinear component of the fit.

I thought that the result should be equivalent to:
tmp2 <- gam(en ~ null, family = quasi(link = identity, variance="mu"))
anova(tmp1, tmp2, test="F")

Analysis of Deviance Table
Response: en

Terms Resid. Df Resid. Dev Test Df Deviance F Value Pr(F)
1 s(null, df = 2) 50.99451 737.3563
2 null 52.00000 860.0650 1 vs. 2 -1.005485 -122.7088 7.6353 0.007856061

Which it clearly isn't.

What is it I have not understood?

Yours gratefully,
Ottar Bjornstad

--------------------------------------------
Ottar N. Bjornstad http://www.uio.no/~ottarnb/
Until 1st of May: NCEAS, 735 State St., Suite 300, Santa Barbara, CA 93101-3351
Tel. + 1 (805) 892 2511, Fax. + 1 (805) 892 2510

After that: Div. Zool., Dept. Biol., Univ. of Oslo, Box 1050 Blindern, N-0316 Oslo, Norway, Tel. + 47 22858370, Fax. + 47 22854605, mailto:ottarnb@bio.uio.no
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news