The more recent versions of SPSS provide heteroskedasticity tests (e.g.,
White's test, modified Breusch-Pagan) as an option when conducting
univariate GLM. Can these tests be used for testing homogeneity of variance
in ANOVA models, and if so, how does the model need to be defined within
SPSS (e.g., 'use predicted values', 'use univariate model', as 'custom')?
References to support this are greatly appreciated. Most texts and
references clearly differentiate these heteroskedastic tests to regression,
however I heard they can be modified to use as a homogeneity tests in ANOVA
but am having a difficult time finding references to indicate such. Thanks
so much.
--
Sent from:
http://spssx-discussion.1045642.n5.nabble.com/=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
