The effects of partial pairing on scale values derived from the method of paired comparisons

Abstract

How can the measurement of attitudes and opinions, by the use of scales involving the method of paired comparisons, be improved? An attempt was made to describe the relationship between scale values derived from partial and complete paired comparison judgments. "One complete and five partial pairing schedules were used, and 60 Ss, who had been randomly assigned to one of these six conditions, scaled the names of 30 nationality groups. The results of the study indicated that partial pairing scale values rather closely approximated those obtained from a complete pairing when the number of observations was reduced as much as 50% of the complete pairing matrix. Beyond this point further reduction seemed to yield a more drastic modification in the scale values obtained. The rating task implications of partial pairing were discussed." (PsycInfo Database Record (c) 2025 APA, all rights reserved)