Change Point Analyses in Correlated Data

• Activity Number: 489
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Kyungduk Ko*
• Companies: Boise State University
• Keywords: Change Point ; Effective Number of Independent Observation ; Levene's Statistic

Abstract

We propose methods to detect a change point in a sequence of correlated data. For test statistics to detect a change point in mean or variance, the ordinary F-statistic is adopted and modified to address correlations among data. A main modification is done by applying the concept of effective number of independent observations (ENIO) to the F-statistic, which is first introduced by Bayley and Hammersley (1946) for removing or reducing the effect of autocorrelation existing in data. Levene's idea is incorporated to the ENIO-modified F-statistic to detect a change point in variance. Simulation study and an application to real data are presented.