Research methods to avoid Bias in categorical ratings of brightness

This article presents evidence of potential contraction bias in the category rating task associated with the stimulus presentation sequence, response range and response range anchors, and a grouping bias associated with the number of stimuli and response categories. These biases tend to reduce the difference between ratings given to stimuli. It is demonstrated that such bias is sufficient to hide differences in brightness under lighting from lamps of different spectral power distribution but that precautions can be taken to successfully counter the bias. Research methods that can be employed to avoid bias in categorical ratings of brightness are summarized.