Showing papers on "Pairwise comparison published in 1983"

The appropriateness of analysis of variance and multiple-comparison procedures.

Abstract: Consider the situation where samples have been obtained randomly from each of five populations A, B, C, D and E. The question arises, 'How should the data be analyzed?'. A common answer to this question is 'Do one-way analysis of variance and, if P < .05, make pairwise comparisons using an appropriate multiple-comparison algorithm'. This is the approach presented in virtually all textbooks, subscribed to by most statisticians, and expected (if not required) by most journals. Conversely, the alternative answer, 'Do 10 two-sample t tests', would typically be considered naive and likely to elicit the response, 'You should consult a statistician'. Perhaps most importantly, medical research is likely to fare poorly among reviewers when the statistical analysis is judged to be naive. Thus the purpose of this

Aggregation (Composition) Schema For Eigenvector Scaling Of Criteria Priorities In Hierarchical Structures.

Abstract: An eigenvector approach to scaling of choice alternatives based upon multiple attributes evaluated on a relative basis by paired comparisons is reviewed. It is shown that neither compensatory aggregation (composition) nor noncompensatory models may be consistent or appropriate for an individual's utility. An alternative approach, where the individual (judge) evaluates all attributes simultaneously, and thereby, performs a subjective aggregation (composition) across all attributes simultaneously, is proposed. Eigenvector scaling may then be used to "prioritize" choice alternatives based upon aggregated pairwise comparisons.

Linear Models Methods for Some Rank Function Analyses of Ordinal Categorical Data

Abstract: This paper is concerned with the analysis of ordinal data through linear models for rank function measures.Primary attention is directed at pairwise Mann-Whitney statistics for which dimension reduction is managed by use of a Bradley-Terry log-linear structure.The nature of linear models for such quantities is contrasted with that for mean ranks (or ridits).Aspects of application are illustrated with an example for which results of other methods are also given.

A Review of Simultaneous Painvise Multiple Comparisons

Abstract: A Review of Simultaneous Pairwise Multiple Comparisons. Simultaneous pairwise comparisons can be accomplished with numerous multiple comparison procedures. The methods differ in two essential ways: the choice of critical value and the specification of the estimated standard error of the mean difference. Those methods that assume homogeneous variances are not robust to violations of this assumption. The methods are contrasted via a numerical example. Results of recent Monte Carlo work are described. A choice between the Games-Howell, Dunnett, and Cochran procedures is recommended.

'Pairwise Balanced' Latin Square Should Always be Used for within-Subjects Designs.

Abstract: : The use of repeated measures designs in many areas of psychological research has prompted concern for the potential confounds inherent in the interpretation of treatments that have been included as within-subject variables. Of the solutions proposed for this problem, the most commonly adopted strategy is the use of Latin Square counterbalancing orders for treatment presentation. Traditional Latin Square designs ensure that each of the experimental treatments included as part of the within-subject factor(s) is administered in each serial position of the treatment sequence. The present paper presents a discussion of a novel technique for the generation of a subset of Latin Squares that control for two additional features that are seen to be important in many research situations, i.e., pairwise priority and distance. Such Latin Squares are referred to as 'pairwise balanced' Latin Squares. The relative advantages of using such Latin Squares in repeated measures designs are discussed. (Author)