PERBANDINGAN METODE DETEKSI PELANGGARAN ASUMSI KLASIK DALAM EKONOMETRIKA MODERN

Abstract

This research discusses a comparison of classical assumption violation detection methods in the context of modern econometrics. In econometrics, classic assumptions, such as homoscedasticity, normality, and the absence of multicollinearity, are often the basis of regression analysis. However, with advances in technology and more complex data, concerns about the validity of these assumptions have increased. This literature study involves an in-depth review of the literature related to detecting violations of classical assumptions in modern econometric regression. Current methods, such as the use of bootstrap techniques, robust tests, and non-parametric approaches, are the main focus in the literature. This research details the strengths and weaknesses of each method, and provides a better understanding of their effectiveness and applicability in addressing the problem of violation of assumptions. The results of the literature study show that classical assumption violation detection methods based on high technology, such as resampling techniques and robust approaches, can provide more reliable results in dealing with the complexity of modern data. It is hoped that an in-depth understanding of this framework can assist econometric researchers and practitioners in choosing the most appropriate method for their specific data conditions, thereby increasing the validity of regression analysis results