정보화정책의 진단과 발전방향 : 선형 구조방정식 모형을 이용한 정보화지수의 계측 = Korean Informatization Policies : Analysis and Recommendations ; A New Way of Constructing an Informatization Index Based on the Linear Structural Equation Model

다국어 초록 (Multilingual Abstract)

We present a new way of constructing informatization indices based on the linear structural equation model. Our model is composed of a set of measurement equations and a set of structural equations: the measurement equations specify the relationships between seven observed indicator variables reflecting the informatization levels in the four major areas of informatization and four unobserved latent variables corresponding to the four sub-indices of informatization while the structural equations specify the interrelationships between four unobserved latent variables. We estimate the resultant model by FIML and we define the estimates for four sub-indices of informatization by Bartlett factor scores for the unobserved latent variables. Even though our study and 『National Informatization White Paper 2002』(National Computerization Agency, 2002) share subcategorization of the informatization index, composition and classification of the observed indicator variables, and the numerical values of the observed indicator variables, their estimates for and inter-country rankings of the informatization sub-indices differ. Within the framework of the linear structural equation model, we explain shortcomings inherent in National Computerization Agency`s (2002) index construction. We also indicate ways to extend our model specification further within the linear structural equation model so as to overcome other shortcomings in existing informatization indices.