dear colleagues;
designing a study to test for reliability a scale measuring behavioural
problems in older people in residential care
if using the bland altman test for comparing differences between two
methods of measurement to test for inter rater reliability; and i have
200 patients and 10 lots of 2 raters assessing, each pair assessing the
same 10 patients, is it right to use the Altman nomogram to assess the
power of the study to be able to detect a difference between the two
raters of at least x (where x is the difference in score judged to be
clinically significant).
If so, doesn't this ignore differences between the 10 pairs of raters;
if not can you help me?
how should we optimise the balance between numbers of subjects vs
numbers of raters to maximise the power of the study to determine inter
rater reliability?
Owen Dempsey
General Practitioner
Senior Research Fellow
LEEDS UK
|