Stata Version 7 question.
I want to compute the intraclass correlation coefficient (ICC,
formula from Bartko), which is based on the mean squares obtained
via repeated measures ANOVA. I do not see in the Stata manual
any scalars etc from the saved results from ANOVA that actually
refer to the SS of person or to the within-subjects factors (both
of which are needed for the Bartko ICC). Can anyone point to a
better understanding of how I might go about doing this? I might
add that I am presently using in an ado file a computational
formula like that which one would use with a hand calculator. It
works fine except that when the vectors are very close (ie, very
tiny discrepancy between them), the result for the ICC exceeds
1.0, and since in ANOVA the sum of the SS for the component parts
never exceed the total SS, it seems that using the ANOVA-based
computation would be consistently more accurate.
thank you
richard ohrbach
--
-------------------------------------------------
Richard Ohrbach, DDS, PhD
Associate Professor
Department of Oral Diagnostic Sciences
University at Buffalo
355 Squire Hall
Buffalo, NY 14214
Voice 716-829-3590
FAX 716-829-3554
-------------------------------------------------
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
