Abstract
The aim of this study is to propose a method to analyze objects in meta-analysis in symbolic data analysis, and represent an application for real data. Meta-analysis aims to obtain a higher level of evidence by integrating multiple studies. Then, it is important to take heterogeneity among the studies into consideration. When we regard the studies as concepts in symbolic data analysis, we can find out factors which may bring heterogeneity. The study proposes a method of hierarchical symbolic clustering when concepts are described as two-way contingency tables. The target data from multiple randomized trials on acupuncture clinical trials contain two-way contingency tables of outcomes and interventions, and individual patient data. The results reveal interpretable clusters and detect some variables for heterogeneity.
