Belief Function Approach to Evidential Reasoning in Causal Maps

The purpose of this chapter is to demonstrate the use of evidential reasoning approach under Dempster-Shafer (D-S) theory of belief functions to analyze revealed causal maps. Revealed causal mapping (RCM) technique, as applied in this chapter, is a qualitative method used to develop or extend understanding of a phenomenon within a specific context. The map can be used to develop models, either as grounded theory or evocative theory building. The example referenced in this study used interview data as the primary source in the RCM method. The participants from information technology (IT) organizations provided the concepts to describe the target phenomenon of Job Satisfaction; they also identified the associations between the concepts. The researchers used coding rules to aggregate similar concepts to produce a composite RCM. The researchers proposed potential evidence measures that could be used to evaluate the model. This chapter discusses the steps necessary to transform a causal map into an evidential diagram. The evidential diagram can then be analyzed using belief functions technique with survey data, thereby extending the research from a discovery and explanation stage to testing and prediction. An example is provided to demonstrate these steps. This chapter also provides the basics of Dempster-Shafer theory of belief functions and a step-by-step description of the propagation process of beliefs in tree like evidential diagrams.