Observer 2
|
Observer 1
|
|
|
diagnosis 1
|
diagnosis 2
|
diagnosis 3
|
subtotal
|
diagnosis 1
|
a
|
b
|
c
|
a + b + c
|
diagnosis 2
|
d
|
e
|
f
|
d + e + f
|
diagnosis 3
|
g
|
h
|
i
|
g + h + i
|
subtotal
|
a + d + g
|
b + e + h
|
c + f + i
|
a + b + c + d + e + f + g + h + i
|
observed agreement as a proportion =
= (a + e + i) / (a + b + c + d + e + f + g + h + i)
expected agreement by chance as a proportion =
= (((a + d + g) * (a + b + c)) + ((b + e + h) * (d + e + f)) + ((c + f + i) * (g + h + i))) / ((a + b + c + d + e + f + g + h + i)^2)
kappa by proportion =
= ((observed agreement as a proportion) – (expected agreement by chance as a proportion)) / (1 – (expected agreement by chance as a proportion))
standard deviation =
= SQRT (((observed agreement) * (1 – (observed agreement))) / (((total number) * ((1 – (expected agreement by chance))^2)))
95% confidence interval =
= (calculated kappa) +/- (1.96 * (standard deviation))
Interpretation:
• minimum value for kappa statistic: < 0
• maximum value: 1
• The higher the number, the greater the level of agreement between the 2 observers.
Result for Kappa
|
Strength of Agreement
|
< 0.00
|
poor
|
0.00 – 0.20
|
slight
|
0.21 – 0.40
|
fair
|
0.41 – 0.60
|
moderate
|
0.61 – 0.80
|
substantial
|
0.81 – 1.00
|
almost perfect
|
from page 165, Landis and Koch (1977)