Trying to calculate the inter-reliability of one rater for 3 different variables which was measured 10 time twice. Each column represents the difference between first and second measurements. Not 100% sure if this is the correct way to calculate the inter-reliability error. I am not sure but maybe I should calculate a Kappa coefficient.
ID Delta_A Delta_B Delta_C
1 300206 -0.1 -0.2 1.3
2 100114 0.1 -0.4 -1.0
3 200211 0.0 -0.2 -1.0
4 200210 0.1 0.1 -0.3
5 200306 -0.1 -0.1 0.9
6 200212 0.0 -0.2 -1.0
7 100128 0.0 0.1 -2.6
8 200317 -0.1 0.0 0.9
9 200126 -0.1 -0.3 -0.3
10 100126 -0.1 -0.6 -0.4
I used icc(df$Delta_A)and I got this error.
Single Score Intraclass Correlation
Model: oneway
Type : consistency
Subjects = 10
Raters = 1
ICC(1) = NA
F-Test, H0: r0 = 0 ; H1: r0 > 0
F(9,0) = NA , p = NA
95%-Confidence Interval for ICC Population Values:
NA < ICC < NA
Warning messages:
1: In qf(1 - alpha/2, ns - 1, ns * (nr - 1)) : NaNs produced
2: In qf(1 - alpha/2, ns * (nr - 1), ns - 1) : NaNs produced
Any help is appreciate.
Thank you-
