Modifications in the format of the mixed standard scale

Blanz and Ghiselli (1972) proposed the mixed standard scale (MSS) as a format for reducing leniency and halo in ratings. They contended that the randomization of scale items to disguise the dimensions prevents raters from distorting their ratings. Another feature of the MSS format that is disguised from raters is its ability to count the number of illogical response patterns committed by a rater. In the present research, the original MSS format was compared to three modifications: (a) a format that did not disguise dimensions; (b) a format that informed raters that their ratings could be evaluated for logic; and (c) a combination of the undisguised and logical formats. The original format and the three modifications along with a behaviorally anchored rating format were used by 833 college students to evaluate their instructors. Analyses indicated no significant differences between the MSS formats in leniency and halo. Further, the MSS formats did not differ in numbers of illogical response patterns. All formats suggested that raters who committed more illogical responses also showed more leniency and halo in their ratings. Finally, a multitrait-multimethod analysis indicated that the MSS formats did not differ in convergent and discriminant validities between themselves and in comparison to the anchored format. In sum, the randomization of items in order to disguise dimensions from the raters is not necessary in order to achieve the desirable properties of the MSS format.