Dear Listmembers, I came across several clinical papers recently where test-retest variability (between 2 sessions) was quantified in RMS terms. My main concern with this is that in these studies, the 'true' value was not known, but was estimated as the mean measurement of the 2 sessions. My questions: shouldn't one quote sample SD instead of RMS (population SD), as otherwise the variability estimate depends on the number of test sessions available for analysis? Any other limitations of this analysis that spring to mind? Any other thoughts or references? Paul ******************************************************************* Paul H Artes | e-mail: [log in to unmask] Department of Ophthalmology | office tel: (902) 473-3240 Dalhousie University | fax: (902) 473-2839 Halifax, Nova Scotia, CANADA | *******************************************************************