5. Methodology of the Empirical Study on ESG and Credit Risk

This chapter delves into the methodology of empirical studies on ESG (Environmental, Social, and Governance) and credit risk, with a focus on linear and logistic regression analysis. The text begins by explaining the fundamental principles of linear regression, including simple and multiple linear regression, and how these models are used to predict dependent variables and uncover cause-effect relationships. It then transitions to logistic regression, detailing its application in predicting categorical or dichotomous variables and analyzing nonlinear relationships. The chapter also covers the assumptions and limitations of these models, as well as the estimation methods used, such as the maximum likelihood method. Additionally, it discusses the interpretation of coefficients and the importance of control variables in enhancing the internal validity of the results. The chapter concludes with a detailed explanation of the sample construction process and the research design used to test the hypotheses. By the end of the chapter, readers will have a comprehensive understanding of the methodological approach to studying the impact of ESG performance on credit risk, equipped with the knowledge to apply these statistical models in their own research.