Title: Parameter Estimation in Panel Data Regression with Fixed Effect Model Using Ordinary Least Square Method
Authors: Adma Novita Sari, Suliyanto, M. Fariz Fadillah Mardianto, Sediono
Volume: 8
Issue: 12
Pages: 89-94
Publication Date: 2024/12/28
Abstract:
Panel data regression combines cross-sectional and time-series data, allowing for in-depth analysis by incorporating individual and temporal dimensions simultaneously. Among various approaches, the Fixed Effect Model (FEM) is widely utilized to control individual-specific heterogeneity that is constant over time. One common method to estimate parameters in FEM is the Ordinary Least Square (OLS) method, which introduces dummy variables for each individual unit to capture unobserved effects. Despite its straightforward implementation, the performance of OLS in handling large datasets or complex data structures, such as heteroskedasticity and serial correlation, remains underexplored. This study aims to analyze the effectiveness of the OLS method in estimating FEM parameters across diverse scenarios, including varying numbers of individuals, time periods, and data complexities. Through a combination of theoretical analysis and simulation experiments, this research evaluates the robustness and accuracy of OLS under different conditions. The findings contribute to understanding the practical applications of OLS in panel data regression, highlighting its strengths and limitations. Moreover, this study provides updated insights into parameter estimation techniques for FEM, ensuring relevance in addressing modern data challenges. These results serve as a guideline for researchers employing FEM in various scientific fields, aiming to enhance estimation accuracy and reliability.