The Structures of Statistical Explanation: Statistical Relevance Condition in Subgroup Analysis of Randomized Clinical Trials

Abstract

The subject of analysis undertaken in the article is the procedure of statistical explanation. For this purpose a specific type of biomedical study, so called subgroup analysis of randomized clinical trials (SA), was discussed. It was shown that the validity of statistical explanations can not be determined by analyzing only the structure of explanation, as is the case in the statistical significance model (S-R) proposed by W. Salmon. Therefore, the statistical explanatory procedure is, in contrast to the other explanatory forms, dependent on the research procedures underlying the formulation of statistical theorems.