IID Residual Test for a Linear Regression Model with Big Data
0
조회
0
다운로드
0
인용
0
참고문헌
초록
Recently, researchers using simple linear regression analysis in very large samples often found that the p-values for testing the slope go quickly to zero as the sample size gets large. In relation, the present study is mainly concerned with such a situation. Consequently, it is shown that additional regression fitting test (IID residual test) can be carried out to address such a p-value problem correctly. For the IID residual test, a reduced U statistic is newly proposed.
Abstract
Recently, researchers using simple linear regression analysis in very large samples often found that the p-values for testing the slope go quickly to zero as the sample size gets large. In relation, the present study is mainly concerned with such a situation. Consequently, it is shown that additional regression fitting test (IID residual test) can be carried out to address such a p-value problem correctly. For the IID residual test, a reduced U statistic is newly proposed.
참고문헌 (4)
- Lin M, "Research commentary-Too big to fail : Large samples and the p-value problem", Inform Syst Res, 24, pp.906~917 (2013)
- 김태윤, "A Different Perspective on the P-value Problem with Big Data", Quantitative Bio-Science, 39(1), pp.71~74 (2020), 자연과학연구소[DOI]
- Cho JC, "Generalized runs tests for the IID hypothesis", J Econometrics, 162, pp.326~344 (2011)
- Kim TY, "Central limit theorems for reduced U-statistics under dependence and their usefulness", Aust Nz J Stat, 55, pp.387~399 (2013)[DOI]