"The coefficient of agreement in determining product preferences": Correction

Abstract

Reports an error in the original article by B. Balinksy, M. L. Blum and S. J. Dutka (Journal of Applied Psychology, 1951[Oct], 35, 348-351. Between the time that galley proofs were approved by the authors and the publication of the above article a number of unfortunate errors in symbols occurred. The correct version is presented here. (The following abstract of this article originally appeared in record 1952-04227-001 .) The coefficient of agreement is most appropriately used in conjunction with the method of paired comparisons. It enables the experimenter to measure the degree and test the significance of agreement among observers as to their preferences for a series of items offered for consideration. Unlike the method of ranks, this technique does not depend upon any assumptions of linearity in the variate being measured. Because of this fact, the method is advisable when subjective times such as preferences for food, clothes, and so on are measured. It can also be readily used in the construction and testing of attitude and opinion scales. In this report the preferences of a jury of observers for four types of ice cream were determined by the paired comparisons method and the results tested by using the coefficient of agreement. (PsycInfo Database Record (c) 2025 APA, all rights reserved)