Using scatterplots to recognize violations for ANOVA tests

[raul...@gmail.com]

Hello, friends. I have been following this group for some time and now
the time has come for my first post. Actually, I am working my way
into statistics and SPSS by myself, and enjoying it!

I want to test the interaction effects of two factors (intensity of
involvement in community service and gender) on one dependent variable
(prosocial action) by means of a two-way ANOVA. Factor 1 has for
levels and factor 2 has two. I want to be sure that the assumptions
necessary for the correct use of ANOVA's are taken care of, and I have
read that a way to recognize violations for ANOVA tests is using
scatterplots. Specifically, I want to see if my distributions are
normal and variances are equal. So, my questions are; how can I make
that scatterplot using SPSS and what I should I pay attention to once
the scatterplot appears in the output?

Thanks a lot.
__

Raul Lozano.

[Kylie]

Hi Raul,

The two main plots that we would assess from an ANOVA are the
residuals versus predicted values for constant variance, and a Q-Q
plot of the residuals to assess normality. The former you can request
as part of SPSS' output ('residual plots'), but for the Q-Q plot you
will need to 'Save' the residuals when you run the ANOVA and then plot
them yourself using the Descriptives > Q-Q plot procedure. This page
is a good SPSS-specific writeup of ANOVA including more information
about the assumptions: http://faculty.chass.ncsu.edu/garson/PA765/anova.htm.

Hope this helps.
Kylie.