Jamshid C. Hosseini

Using the Analytic Hierarchy Process as a Two-phase Integrated Decision Approach for Large Nominal Groups

When decision makers who comprise a large nominal group face an unstructured decision problem and no simultaneous interactive communications are available, problem identification and consensus building are difficult, if not impossible. Few tools are available to assist decision makers in this situation. The Analytic Hierarchy Process (AHP) has typically been used to evaluate a set of alternatives after a decision problem has been structured as a hierarchy with various levels of criteria above the alternatives. With a group of decision makers, AHP has been used to evaluate those alternatives either by consensus building or by combining judgments or priorities using the geometric mean to aggregate their preferences. In this paper, we extend the use of AHP to a situation involving a large nominal group of dispersed decision makers where the entire hierarchy is not defined at the outset. In particular, we use the AHP as an integrative approach to identify the priorities of the various criteria and then use those priorities to screen and consolidate a large set of potential alternatives. This results in considering a reduced set of alternatives that will be affected by the more important criteria. The consolidated set of alternatives is evaluated by each individual in the group using AHP, combined using the geometric mean, and the results are synthesized to obtain the overall priorities of the alternatives. The approach is demonstrated and evaluated in a case study to select an alunmi anniversary gift to the U.S. Coast Guard Academy with a large nominal group of decision-makers dispersed throughout the United States.

The two-group discriminant problem with equal group mean vectors: An experimental evaluation of six linear/nonlinear programming formulations

Abstract Various parametric and nonparametric approaches to multiple discriminant analysis attempt to discriminate among or classify entities (e.g., loan applicants, customers, employees, businesses) based on several of their distinguishing characteristics called discriminant variables. Statistical parametric procedures require that the mean vectors of discriminant variables for the populations of entities be different across groups. This requirement may not always be met in practical settings. This paper reports on a preliminary Monte Carlo simulation experiment which compares the performance of six 1 p -norm distance models including two linear and four nonlinear formulations as well as two statistical procedures to address the discriminant problem under equal mean vectors. The experimental data were generated from multivariate normal or nonnormal populations with equal or unequal dispersion matrices and with or without outliers. The results indicate that, when population mean vectors are equal, the most significant characteristics which affects the performance of all of the methods is the similarity (or dissimilarity) of dispersion matrices. The departure from normality and the presence of outliers and some of the interactions between these three factors are also statistically significant

A FRAMEWORK FOR ANALYZING PERCEPTUAL QUALITY IN MANUFACTURING

During the last two decades, manufacturing managers implemented programs such as Total Quality Management and Quality Function Deployment to meet customer demands. Many of these programs have been unsuccessful. Our framework addresses the major sources ..