Incomplete Block Rank Analysis: On the Appropriateness of the Model for a Method of Paired Comparisons

Abstract

A procedure is developed for testing the appropriateness of a model previously introduced for the rank analysis of incomplete block expts. based on paired comparisons. The procedure is applied to three expts. in which judges were required to assess taste, preference and appearance. Twenty chi-square tests of goodness of fit made at the 5% level yielded only one significant result. This is interpreted as giving some assurance that the model is valid, although the primary purposes of the paper are to present the tests and to illustrate their application.