Alan T. Arnholt (arnholt@math.appstate.edu)
Mon, 8 Nov 1999 10:34:17 -0500 (EST)
Dear R list,
What is the proper way to specify a nested model so that
the F values agree with the expected mean square errors?
Specifically, suppose I have a design where "Heads" are
nested within "Machines". I would like to model the
following Y_ijk = Mu + Machine_i +Head_j(i) +Error_k(ij).
Using the commands below,
> summary(aov(Strain~Machine + Head%in%Machine ))
Df Sum Sq Mean Sq F value Pr(>F)
Machine 4 45.07 11.27 1.0532 0.38762
Machine:Head 15 282.87 18.86 1.7625 0.06252 .
Residuals 60 642.00 10.70
the F value to test "Machine" uses the MS residual
instead of MS Head(Machine).
When I use the following no F value is reported
> summary(aov(Strain~Machine + Head%in%Machine + Error(Machine)))
Error: Machine
Df Sum Sq Mean Sq
Machine 4 45.075 11.269
Error: Within
Df Sum Sq Mean Sq F value Pr(>F)
Machine:Head 15 282.88 18.86 1.7625 0.06252 .
Residuals 60 642.00 10.70
Is the second approach the best I can do?
Thanks ALAN
___________________________________
Alan T. Arnholt
Office phone: (828) 262 - 2863
Office fax : (828) 265 - 8617
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
This archive was generated by hypermail 2.0b3 on Tue 04 Jan 2000 - 13:34:00 EST