ABSTRACT
This chapter describes the many ways people have identified causes in specific cases, from enumerating associations, to conducting experiments, to comparing model fits. Each method has strengths and limitations, which are better understood by reviewing their origins. Each of the methods has informed our overall approach by suggesting different ways that evidence of a causal relationship can be derived, synthesized, and compared. Evidence is synthesized using qualitative weights so that all relevant types of evidence can be included. The candidate causes are compared with respect to the full body of evidence so that the cause or causes that are best supported can be identified. Documenting the occurrence of an association between a candidate cause and an effect may simply involve co-occurrence in the single instance of an effect or the enumeration of instances in which the two are associated. Diagnosis is the identification of a cause by recognizing characteristic signs and symptoms.
