The Measurement and Assessment of Colour Differences for Industrial Use. IV—The Accuracy of Visual Assessments

The magnitude of the errors in the visual assessment of colour differences is discussed. The errors are likely to vary with the particular form of visual assessment. Methods are described for estimating visual errors from data in which the visual assessments are expressed as percentage acceptance and rank orders as well as ratios of other colour differences. The results indicate that experienced observers tend to have a greater degree of discrimination than inexperienced observers. Between‐observer variation is greater than within‐observer variation, suggesting that the mean of many repeat observations by a single observer may differ significantly from the mean derived from a number of different observers. For colour passing, the error of a single observer varies with the size of the colour difference and is given by the equation: σ=0.4+ 0.25 δν where δν is the true visual difference expressed in units equivalent to the average unit given by the AN40 equation.