Skip to main content
. Author manuscript; available in PMC: 2018 Mar 1.
Published in final edited form as: Stat Methods Med Res. 2016 May 16;27(3):812–831. doi: 10.1177/0962280216643347

Table 7.

Results for the beam mammogram study9 where 104 radiologists (J =104) classified mammograms of 148 patients (I = 148) using an ordered BIRADS scale with C = 5 categories (1 = normal; 2=benign; 3=probably benign; 4=possibly malignant; 5 = probably malignant).

Parameter Symbol Estimate S.E. Z-value
Ordinal GLMM:
Thresholds: (α0 = −∞, α5 = +∞)
 Between categories 1 and 2 α1 −0.897 0.135 −6.643
 Between categories 2 and 3 α2 −0.197 0.135 −1.460
 Between categories 3 and 4 α3   0.761 0.135   5.630
 Between categories 4 and 5 α4   2.539 0.137 18.574
Subject random effect variance
σu2
  2.442 0.427
Rater random effect variance
σv2
  0.158 0.073
Rho ρ   0.678 0.026
GLMM-based observed agreement p0   0.430
GLMM-based observed association (quadratic weights) p0a   0.907
Measures of agreement:
 Model-based (unweighted) kappa
 (Nelson and Edwards41)
κm   0.241 0.015
 Fleiss’ kappa (Fleiss47) κF   0.297 0.001
 Light and Conger’s kappa
 (Light,48 Conger49)
κLC   0.298
Measures of association:
 (with quadratic weights)
 Model-based weighted kappa κma   0.475 0.022
 Shrout and Fleiss’ ICC[2,1]
 (Shrout and Fleiss28)
  0.652 95% c.i. = (0.601, 0.706)
 Cohen’s GLMM-based weighted kappa κGLMM,a   0.611