Re: Main Effects and Interactions

[Karl L. Wuensch]

    Here is SAS GLM output for a 2x2 ANOVA, unequal cell sizes, type III sums of squares.  First the full factorial design:

 

Source        DF         Type III SS

deattr         1       1286.325649

gender         1          3.955131

deattr*gender  1         15.873457

 

    Now I drop the two main effects and run the analysis again:

Source           DF        Type III SS

deattr*gender     3        1312.292621

 

    As you can, both the sums of squares and the df previously assigned to the main effects are now assigned to the interaction term.

----- Original Message -----

From: "Arthur Tabachneck" <[EMAIL PROTECTED]>

To: <[EMAIL PROTECTED]>

Sent: Friday, February 13, 2004 8:10 PM

Subject: [edstat] Main Effects and Interactions

Is it legal to test for an interaction without testing for the main effects
of the variables included in those interactions.  And, in the case it is
(which I recall it isn't), does one still have to account for the degrees of
freedom used by the non-tested main effects?

Hoping that was sufficiently clear,
Art


.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================