Experimental comparison of brightness judgments obtained by rating and bisection methods

The consistency of psychophysical scales obtained by different measurement procedures is an important topic for research. Two experiments tested whether the scales obtained by the rating and bisection methods were consistent. 40 university students bisected intervals on the brightness continuum and subsequently rated the brightnesses determined by such bisections, or they first rated brightnesses and then bisected brightness intervals. Analysis shows the scales obtained by these methods were mutually consistent when a large number of response categories (the integers from 0 to 100) were used for ratings. When a small number of response categories (the integers from 0 to 6) were used, the methods were consistent only when ratings were produced after the bisections. This suggests that bisections may have increased the consistency of the rating scale by generating an internal representation of equally spaced sensory magnitudes subsequently used for ratings.