이성우 안동환 지우석 2018-11-15T08:47:13Z 2018-11-15T08:47:13Z 2004-06 JRD26-2-3 http://repository.krei.re.kr/handle/2018.oak/18056 The purpose of the present study is to investigate the analytical structure of the multi-level model, which is also called to hierarchical model, random coefficient model etc. Although interests on the multi-level model are recently increasing in natural sciences as well as social sciences, the accurate understanding of the analytical structure of the model has never been dealt with. This analytical paper constructs five analytical questions and derives diverse ways to reach the answers for each question. The present study also argues that the multi-level model provides a powerful tool for analysis of the data within which the number of observations varies. The purpose of the present study is to investigate the analytical structure of the multi-level model, which is also called to hierarchical model, random coefficient model etc. Although interests on the multi-level model are recently increasing in natural sciences as well as social sciences, the accurate understanding of the analytical structure of the model has never been dealt with. This analytical paper constructs five analytical questions and derives diverse ways to reach the answers for each question. The present study also argues that the multi-level model provides a powerful tool for analysis of the data within which the number of observations varies. I. Introduction II. Theoretical Concerns III. General Structure of Multi-level Model Ⅳ. Distributions in Known Covariance Matrices V. Distributions in Unknown Covariance Matrices Ⅵ. Methods for Testing Some Null Hypotheses Ⅶ. Solution for Heteroscedasticity of Unknown Form Ⅷ. Calculation of Upper-level Expectation in Logit Model Ⅸ. Conclusions 서울대학교 Analtyical structure of the multi-level model Analtyical structure of the multi-level model KREI 논문 66 41 Lee, Sungwoo An, Donghwan Zhee, Woosuk page. 41 - 66 Multi-level Model Random Coefficient Covariance Matrix Heteroscedasticity Multi-level Model Random Coefficient Covariance Matrix Heteroscedasticity