|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
DIFF-test and cluster sampling |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
|
Dear Dr. Muthen, For my SEM-analyses, I applied cluster sampling because of non-independence of the participating adolescents (i.e., in most cases adolescents were part of the same family). Consequently, I tried to ascertain no differences between these clustered adolescents (i.e., the older adolescents versus the younger adolescents). However, is it possible to do a DIFF-test while using cluster sampling? In other words: Can I do a DIFF-test for non-independent groups? Yours sincerely, Rebecca de Leeuw |
|
|
You should not do a multiple group analysis when the groups are not independent. You say you applied cluster sampling. What do you mean by that? If you want to take the non-independence into account, you can use TYPE=COMPLEX; or TYPE=TWOLEVEL; |
|
|
Thank you, for your message. For taking into account the non-independence I clustered the adolescents from one family (CLUSTER IS) while using TYPE=COMPLEX. Yours sincerely, Rebecca de Leeuw |
|
|
Sounds good. |
|
Back to top |
|
|