|
|
Hello,
I have a scale with more items (28) rated by more people and I want to
compute the rwg(1) and rwg(j) for it. I want to do the calculations myself. From
what I understand so far, the formula for rwg(1) includes the observed variance
among raters for an item divided by the expected distribution in the case of no agreement. This will be
substracted from 1 to result the rwg(1) for an item. And this continues for
every item in the scale to compute their rwg (1). An average rwg(1) would be
obtained by averaging individual item rwg(1).
The formula for
rwg(j) containes the mean of the observed variances on J items. Would this mean
that in order to obtain the rwg(j) I would use the average rwg(1) obtained in
previous calculations or if not what variances will I use in this calculation? I
now there is a syntax but I prefer this time the "pen and paper" option. Can
anyone help me with some advice?
Thank you!
Andra Toader
E-mail: [hidden email]
|
|
Locked
