32 – Modeling Single and Multiple Time Series
Temporal Aggregation Effects on a Structural Mean-Change of Time Series
Sponsor: Business and Economic Statistics Section
Keywords: Temporal Aggregation, Mean Change, Likelihood Ratio Test, Cumulative Sum Test
Bu Hyoung Lee
Temple University
William W. S Wei
Temple University
In this article we investigate the effects of temporal aggregation on a mean change of time series, through the two statistical tests-the likelihood ratio (LR) test and the cumulative sum (CUSUM) test to detect the mean change. We propose a modified LR test statistic when aggregate data are used for testing. Also we show that the CUSUM test statistic is free from the temporal aggregation effect. The Monte Carlo simulations verify the theoretical results.