Abstract
Two techniques are presented for obtaining complete rankings of relatively large numbers of items with minimum difficulty for the respondent. The alternating extremes technique involves the successive selection of items with high and low extreme ranks, in effect ranking inward from the ends of the ordering toward the middle. The successive halvings technique involves the division of the set of items into "high" and "low" halves, followed by the division of the set of items into "high" and "low" halves, followed by the division of each half into such halves, and continuing until all items are ranked. These techniques are compared with a distance scale technique previously suggested in the literature. The results of the test comparisons indicate that the extremes technique is generally preferable to the other two. It exceeds the halvings technique and distance scale in efficiency and, with halvings, produces fewer errors and ties than distance scale. It has considerable "face validity." The distance scale tends to receive higher subjective ratings on accuracy, very possibly because it demands fewer fine discriminations.
Get full access to this article
View all access options for this article.
