Message/Author |
|
Anonymous posted on Friday, January 04, 2013 - 8:43 am
|
|
|
I am looking for some (relatively recent) guidelines/literature on the use of the WRMR as part of the process for assessing model fit in multigroup models. I note that values of <1 are often regarded as indicative of satisfactory fit, however in one of your handouts regarding CFA testing there is a comment that the WRMR is not good with multigroup models. Hence I am looking for clarification on the matter. Thank you |
|
|
I would not use WRMR. It is an experimental fit statistic that has not been well studied. As you mention, we have found it does not work well in a multiple group setting. |
|
Anonymous posted on Saturday, January 05, 2013 - 10:34 am
|
|
|
Thank you. Do you know of any references I can cite regarding use of WRMR in multiple group settings? |
|
|
I don't think Yu's dissertation covered that and I don't recall seeing it discussed. |
|
Back to top |