On Reflection Causality in Medicine in JSM Method of Automated Research Support (Evidence from a Study of Establishing a Link between Negative Schizophrenia and Genetic Parameters)

Abstract

In this paper, we consider some approaches to causality in epistemology and their connection with JSM method of automated research support (JSM method ARS). The JSM method ARS is able to detect various types of cause-and-effect dependencies in data using appropriate strategies. Some approaches to causality in medicine (uniqueness of a cause, multiple causes, and independent causes) and their reflection in JSM method ARS are presented. Using a research of establishing a link between negative schizophrenia and genetic parameters as a case study, it is shown the choice of an appropriate type of causality determined in JSM method ARS using strategies. In the research under consideration, two strategies of JSM method ARS were used: one with the method of simple similarity with a ban on counterexample and the one with the generalized method. The application of the first method demonstrated a low rate of explainability of the initial data array by the generated hypotheses, which does not satisfy the criterion of abductive acceptance of hypotheses. The application of the generalized method improved the explanatory rates, which may indicate a more successful representation of the relationship between genetic parameters and the disease studied in the research. This representation is associated with the ternary similarity relation underlying the generalized method: “cause—set of brakes (blockers) of the cause—effect.”