Date of Award


Document Type

Open Access Master's Thesis

Degree Name

Master of Science in Statistics (MS)

Administrative Home Department

Department of Mathematical Sciences

Advisor 1

Yeonwoo Rho

Committee Member 1

Qiuying Sha

Committee Member 2

Kui Zhang


Rolling window is a popular tool in time series analysis. When conducting hypothesis testing on each window simultaneously, multiple testing problem occurs. In the literature in rolling window analysis, it appears that bootstrap is the most frequently used, if not only, method to address the multiple testing issue. This thesis aims to adapt multiple testing correction methods that are popular in genome-wide association study to the time series rolling window context. In particular, some of these methods require the knowledge of the correlation structure of test statistics. In genetics, this structure can be obtained from an external source, which may not exist in time series. To overcome this difficulty, we adopt the AR sieve idea, which enables the computation of correlation structure based on the estimated AR coefficients. We also present the finite sample simulation to illustrate the performance of these methods.