Hi Stephen
Brown-Forsythe and Welch tests are indeed NOT HOV tests. If you want to
check the HOV assumption, use Levene test. Welch test should be used as
a replacement of the ordinary F test of the ANOVA table if HOV
assumption is not met and the design is unbalanced, like in your case.
HTH,
Marta GG
Salbod, Mr. Stephen wrote:
>
> Good Day Friends,
>
> When should the two options on the one-way procedure: Brown-Forythe
> and Welch be checked? I have three groups of data that have unequal ns
> and unequal variances. I’m planning to use Dunnett C for post-hoc
> analyses. When I checked Welch procedure, PASW Statistics generated a
> pivot table labeled: “Robust Tests of Equality of Means”. Does this
> mean that these are not HOV tests?
>
> Any light you can shine on this issue will be greatly appreciated.
>
> Stephen Salbod, Pace University, NYC
>
--
For miscellaneous SPSS related statistical stuff, visit:
http://gjyp.nl/marta/=====================
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
