This paper studies the change point detection problem under the non-central Student t distribution proposed by Hasan et al. [20, 21]. We compare the performance of the Modified Information-Based procedure with the Bayesian Information-Based approach, highlighting how the former avoids redundant change points by increasing the penalty or complexity term near the beginning and end of the time series. Unlike traditional approaches that rely on the asymptotic distribution of test statistics, both methods provide a practical alternative to detect structural changes. Through simulation studies, we evaluated the performance and consistency of both methods. The stock market prices of Apple Inc. were used to illustrate the effectiveness of the non-central skewed Student t distribution in identifying structural shifts in financial data.
