The Selection of Advance Statistics for Research of Public Administration: A Case Study of Structural Equation Model

Abstract

The research for Public Administration is 5imilar to general social science research. In fact, there are many variables involved in research, meaning many variables especifically causal and consequential research. Multivariate analysis techniques are widely used in research. Logical Regression Analysis, Cluster Analysis , Discriminant Analysis and many other analysis techniques. The above techniques are considered as the most effective tools for data analysis. It is also Very popular for research in doctoral dissertation or graduate research. But there are also disadvantages that the technique can detect a relationship or cause only one relationship at a time. There are only 1 variable per variable. Some statistical scholars call it primitive, which must analyze the pairs of relationships, or called relationship analysis that is the analysis of each variables interested pair, it will cause a lot of errors or fragments of analysis. Therefore, it is necessary to apply advanced statistical techniques to the analysis process by analysis of structural equations model which contains a latent variable and the indicators.