One hundred and twenty-eight subjects responded to one of four differently labeled five-point Likert-type response scales. Although no significant differences in test means and in reliability estimates were found among response scales, the scales did differ on measures of variability. (Author/LMO)

Using both the Behaviorally Anchored Rating Scales (BARS) and the Purdue University Scales, 727 undergraduates rated 32 instructors. The BARS had less halo effect, more leniency error, and lower interrater reliability. Both formats were valid. The two tests did not differ in rate discrimination or susceptibility to rating bias. (Author/GDC)