Application of multiple testing methods in panel unit root testing
| Project/Area Number |
15K03403
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Economic statistics
|
|---|
| Research Institution | Osaka Gakuin University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2015-04-01 – 2018-03-31
|
| Project Status |
Completed (Fiscal Year 2017)
|
| Budget Amount *help |
¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
Fiscal Year 2017: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2016: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
|
|---|
| Keywords | 非定常パネルデータ / 単位根 / 共和分 / 多重検定 / 経済構造変化 / 非線型性 / 共変量 / 定常共変数 / 時系列 / 定常covariate |
|---|
| Outline of Final Research Achievements |
In panel unit root testing, two crucial issues often arise: (1) Impossibility to identify which is the stationary series through commonly used panel unit root tests when the joint unit root null is rejected; (2) multiplicity problem (that is, over-rejection of the null in the repeated application of individual tests). To address these issues, this study employs multiple testing in nonstationary panel settings.
Specifically, this study focuses on three points: (1) the test performance is investigated under various circumstances in actual data analysis. (2) In conducting stepwise multiple testing, the data associated with the null hypothesis rejected in the first step is re-utilized in later steps as a stationary covariate that correlates with other series. The gain in test power is confirmed. (3) Then, some multiple testing methods are employed to analyze actual economic data.
As a result, all these analyses show a good practical performance of multiple testing.
|
Report
(4 results)
Research Products
(17 results)
