Seite 1 von 1

Unrealistically high ICCs

BeitragVerfasst: Do 2. Mär 2023, 20:10
von pamplona
Dear all,

i am calculating reliability in SPSS using ICCs just like in this tutorial: https://www.youtube.com/watch?v=v2d6eyYoh4M
However, the output seems unrealistically high to me (two-way random, absolute): 0,913 (average)

Can you explain what I am doing wrong?

I used this set of data (every column is a different observer, every row a different sample that was measured). The goal is to analyze ICCs between observers:
27 51 17 26 34 45 29
47 52 32 20 52 69 71
57 45 32 44 97 57 50
42 20 27 34 32 45 33
25 12 17 17 17 23 26
7 52 17 7 7 7 15
7 12 12 7 7 7 9
20 12 12 7 7 26 25
35 30 37 19 34 35 33
7 12 7 7 7 12 10
12 12 17 11 9 12 10
27 18 22 30 27 29 20
12 52 17 7 25 18 23
57 52 27 38 27 52 43
22 30 17 16 18 33 25
7 12 17 10 7 10 19
28 13 12 7 12 37 23
11 12 12 7 7 7 10
27 29 22 20 29 35 42
42 20 22 41 15 47 40

Re: Unrealistically high ICCs

BeitragVerfasst: Fr 3. Mär 2023, 07:13
von strukturmarionette
Hi,

- Probably you should restructure your SPSS-data file first

Greetings
S.

Re: Unrealistically high ICCs

BeitragVerfasst: Fr 3. Mär 2023, 10:59
von pamplona
Thanks, in what way?