Skip to main content
. 2016 Jan 13;115(4):1932–1945. doi: 10.1152/jn.00318.2015

Table B4.

Conventional confidence metrics for human and simulated data

BS REL RES UNC Deviance Binary Deviance CSD
Subject 1 0.150 (0.015) 0.081 (0.013) 0.181 (0.016) 0.249 (0.001) 92.6 (8.7) 789.0 (40.8)
Subject 2 0.127 (0.021) 0.074 (0.017) 0.192 (0.016) 0.246 (0.003) 75.3 (10.1) 752.6 (44.8)
Subject 3 0.145 (0.020) 0.081 (0.011) 0.185 (0.019) 0.249 (0.001) 93.6 (9.5) 791.3 (31.0)
Subject 4 0.147 (0.025) 0.076 (0.014) 0.176 (0.012) 0.247 (0.004) 79.2 (12.8) 821.0 (20.6)
Fig. 8, A, E, and I 0.135 (0.019) 0.095 (0.015) 0.207 (0.015) 0.247 (0.004) 79.3 (8.0) 764.6 (37.4)
Fig. 8, B, F, and J 0.141 (0.015) 0.091 (0.012) 0.198 (0.016) 0.247 (0.004) 79.3 (8.0) 807.9 (18.2)
Fig. 8, C, G, and K 0.152 (0.015) 0.093 (0.013) 0.188 (0.017) 0.247 (0.004) 79.3 (8.0) 830.7 (22.7)
Fig. 8, D, H, and L 0.150 (0.014) 0.092 (0.013) 0.190 (0.017) 0.247 (0.004) 79.3 (8.0) 788.0 (21.7)

Brier Score (BS) and its 3 decomposed components: reliability (REL), resolution (RES), and uncertainty (UNC) are shown. For definitions and descriptions, see Brier 1950; Murphy 1973, 1972. We use the formulation commonly used today, which results by dividing Brier's original formulation by 2. Deviance for each of 2 fits are also shown. Mean (SD) across 6 trials for each subject and across 10,000 simulations for each simulated data set are provided.