Table 2.

Variance components for video scoring^*

Polyp Burden Measure†	Subject-to-Subject variance (relative variability)	Reviewer-to-Reviewer variance (relative variability)	Residual variance (relative variability)
ES	797 (0.82)	14 (0.01)	162 (0.17)
MR	1689 (0.81)	22 (0.01)	387 (0.18)
BR	975 (0.81)	18 (0.01)	215 (0.18)
TC	243 (0.78)	14 (0.05)	53 (0.17)
TC2	97 (0.72)	3 (0.02)	34 (0.25)

^*Variance components from linear mixed model analysis done on the various polyp burden measures.

The model included variance components for subject-to-subject variance (biological variability), reviewer-to-reviewer variance (systematic reviewer variability) and residual variance (non-systematic reviewer variability). The table presents restricted maximum likelihood (REML) estimates of the variance components along with relative proportion of variability at that level in parenthesis. High interrater reliability is indicated by high proportions of variability coming from the natural subject-to-subject variability relative to the technical reviewer-related sources.

^†See Methods section for the formulas for calculating the following polyp burden measures: equally spaced weighted counts (ES), mid-range weighted counts (MR), and bottom-range weighted counts (BR). TC = sum of polyp counts in all size categories; TC2 = sum of polyp counts in the 2–4 mm and >4 mm categories.