Collecting and Analyzing Expert Group Judgment Data
Presents methods for collecting and analyzing judgments from groups of experts. The paper addresses issues associated with resolving the procedural and administrative problems involved in selecting a panel of experts, in eliciting informed judgments about the degree of technological advance on relevant projects, and in designing a survey questionnaire for measuring those judgments. Three methods of multivariate analysis are described for quantifying and analyzing group judgment data collected from a panel of experts. Those of the methods that are known in earlier literature include multidimensional scaling of individual differences, and subjective probability procedures, including the Bayesian approach. A new procedure, which was developed specifically for this application, involves the use of simultaneous equation system models in which the response (dependent) variables are categorical and unordered.