基于高层次结构数据的多水平发展模型设计及应用

Design and Application Based on the High-level Structure Data of Multilevel Development Model

研究目标:多水平模型是经济计量分析的一种重要工具,通过对高层次结构数据进行多水平发展模型的构建研究,以期探索嵌套数据的高层次模型建构理论和方法。研究方法:综述了多水平模型研究现状,构建了高层次数据结构发展模型,对比分析了四、三、二水平发展模型的拟合效果,实证分析了农户收入嵌套数据。研究发现:发现了四水平发展模型参数估计显著性更好,模型拟合效果更优。研究创新:凸显高层次嵌套结构数据的多水平发展模型的优势和特点。研究价值:为经济社会领域复杂数据定量化分析提供了一种有效的建模方法。

Objectives： Multilevel model plays an important role in econometric analysis. Through the research on the statistical modeling of multilevel development model for analyzing high-level structural data, it expects to explore the theory and method of high-level hier-archical model establishing for nested structure data set. Research Methods： This paper summarizes the current research status about multilevel model, and then presents a high-level data structure of multilevel development model. Furthermore, we study model fitting effect by comparing the four, three, two level development models and the corresponding estimation results. At the same time, the paper empirically analyzes the nested data based on the farmers＇ income in western area of China. Research Findings： The results show that the significanees of parameters estimation for four level development model are better, and its model fitting effect is even more excellent. Research Innovations： It highlights that what are the advantages to establish a best statistical model for high-level hierarchical structural data set, and how to obtain useful results in applications. Research Value： We provide an effective sta-tistical modeling method of quantitative research for high-level complex structure data in socioeconomic fields.