Hello
I am currently considering the calculation of an intra-class correlation coefficient (ICC) for a two-way mixed model in which the data are on a Likert scale (from 1 to 6). I would therefore be interested to receive suggestions as to the most suitable formula(e) for performing an associated sample size calculation based on statistical power. I am particularly concerned that the assumptions for the sample size calculation should be in agreement with the sort of data that I have, as I am already aware that there are a variety of formulae for predicting sample size calculations for ICCs based on assumptions relating to Normality.
I look forward to receiving some useful suggestions.
Many thanks in advance for your kind consideration of this request.
Margaret
---------------------------------
Inbox full of spam? Get leading spam protection and 1GB storage with All New Yahoo! Mail.
|