Statistical hypothesis testing in growth curve model with missing values

2023 Fiscal Year Final Research Report

• Project/Area Number: 19K20225
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 60030:Statistical science-related
• Research Institution: Tokyo University of Science
• Principal Investigator: Yagi Ayaka 東京理科大学, 理学部第一部応用数学科, 講師 (40823547)
• Project Period (FY): 2019-04-01 – 2024-03-31
• Keywords: 単調欠測データ / 最尤推定量

Outline of Final Research Achievements

Regarding the statistical hypothesis testing problem in growth curve model with missing values, we have obtained results on (1) parameter estimation and testing in growth curve model with monotone missing data.
In addition, related to issue (1), we have also obtained research results on (2) a new testing procedure on mean vectors with monotone missing data, (3) tests on variance-covariance matrices with monotone missing data, and (4) simultaneous testing of mean vectors and variance-covariance matrices with monotone missing data.

Free Research Field

数理統計学

Academic Significance and Societal Importance of the Research Achievements

成長曲線モデルは経時的な実データのモデル化のひとつであり，例えば，人の成長に関して経時的に測定されたデータのモデル化に適している．このようなデータには欠測値が含まれることが多々あり，欠測のパターンが単調（一度欠測が生じるとそれ以降も欠測する）になることもしばしば起こるため，その場合の推定や検定問題などの統計解析手法を開発することは有用であると思われる．