2022 Fiscal Year Final Research Report
Research on model misspecification for longitudinal data analysis
| Project/Area Number |
19K11849
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 60030:Statistical science-related
|
|---|
| Research Institution | University of Tsukuba |
|---|
Principal Investigator |
Maruo Kazushi 筑波大学, 医学医療系, 准教授 (10777999)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
石井 亮太 筑波大学, 医学医療系, 助教 (40835633)
|
|---|
| Project Period (FY) |
2019-04-01 – 2023-03-31
|
| Keywords | モデル誤特定 / ロバストネス / 歪んだデータ / 欠測 |
|---|
| Outline of Final Research Achievements |
In the analysis of data from randomized controlled trials in which outcomes are measured longitudinally, we investigated the effects of misspecification of statistical models and developed robust models that are less sensitive to misspecification. Specifically, we (1) developed a program package for a model that fits well when the shape of the outcome distribution is skewed, and (2) evaluated the robustness of robust variance that allows model misspecification under missing data when estimating the precision of the treatment effect. These research results were published in a peer-reviewed international journal.
|
| Free Research Field |
医学統計学
|
| Academic Significance and Societal Importance of the Research Achievements |
概要における①歪んだ経時データの解析方法のプログラムパッケージの開発について，この方法は当該の状況において新規治療の治療効果の検出力を高めることが示されており，有用な治療のより効率的な開発に寄与することが期待される．②ロバスト分散の有限欠測データにおけるロバストネスの評価について，ある程度どのような状況でも用いることができる分散推定量は解析者にとって非常に便利であり，この性質を明らかにしたことは統計ユーザーにとって有用であると考えられる．
|
