Skip to main content

Table 2 Predictors of reviewer bias.

From: Computer-aided assessment of diagnostic images for epidemiological research

Variable

Fixed parameter estimate*

Standard error

P

Intercept

0.25

0.31

0.417

More opacities

-0.21

0.08

0.005

Darker opacities

-0.07

0.10

0.472

Wide opacities

-0.17

0.10

0.075

Long opacities

0.31

0.10

0.002

Reviewer

0.04

0.27

0.879

  1. *Bias = estimated severity - true severity
  2. In cataract severity score units
  3. Parameter estimates for the fixed-effects from the linear mixed effects model analysis of image factors affecting bias.
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us