2014 Fiscal Year Final Research Report
Research for causal analysis between body mass index and death
Project/Area Number |
24700278
|
Research Category |
Grant-in-Aid for Young Scientists (B)
|
Allocation Type | Multi-year Fund |
Research Field |
Statistical science
|
Research Institution | Yokohama City University |
Principal Investigator |
|
Project Period (FY) |
2012-04-01 – 2015-03-31
|
Keywords | 因果推論 / 疫学 |
Outline of Final Research Achievements |
For the purpose of examining a causal association between body mass index and death appropriately, I developed and examined statistical methodologies when there exists time-dependent confounding. I developed model selection criterion for two kinds of causal models, Marginal Structural Models and Structural Mean Models. I showed by simulation experiments that the proposed criterion can choose the right causal model with the probability that is higher than existing methods. I further developed estimation and sensitivity analysis methods for evaluating the direct and indirect effects of an exposure variable, in the presence of time-dependent confounding. By an analysis of a cohort data, I found that there is not large time-dependent confounding.
|
Free Research Field |
統計科学
|