The models for multi-way contingency tables and exact test using economic data

2021 Fiscal Year Final Research Report

• Project/Area Number: 19K14600
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 12040:Applied mathematics and statistics-related
• Research Institution: Nihon University
• Principal Investigator: IKI Kiyotaka 日本大学, 経済学部, 准教授 (30711593)
• Project Period (FY): 2019-04-01 – 2022-03-31
• Keywords: 数理統計学 / 多元分割表 / 対称性 / リスク資産データ

Outline of Final Research Achievements

In this study, we defined the sevens kinds of model of symmetry for a 3 * 3 * 3 * 3 contingency table. In order to investigate the goodness-of-fit the proposed model, we extracted the changes in the stock prices of companies listed on the First Section of the Tokyo Stock Exchange as analysis data. Focusing on the stock price in August 2009 and selecting the best fitting model by AIC, it was interpreted that the percentage of companies with a higher stock price in the 4th week than the stock price in the 1st week is equivalent to the percentage of companies with similar low stock prices. In other words, it can be seen that the ratio of companies whose stock prices have risen and those whose stock prices have fallen is equal in this month.

Free Research Field

統計科学

Academic Significance and Societal Importance of the Research Achievements

本研究では多元分割表に対して、数学的な妥当性はもちろんのこと、解釈のしやすさを念頭においてモデルの考案を行った。また、連続値を「意味のある差」に基づいていくつかのカテゴリに分類することで分割表を作成することを試みた。株価のデータを分割表に変換し解析を行うことで、分割表解析の観点から、新たな解釈を与えることに成功した。提案手法は経済データのみならず、医学、スポーツ科学、アンケート調査といった様々な分野に応用が可能である。