Statistical Models Of Multilevel And Cross Classification Accumulated Method

In recent years, the multilevel statistical analysis model has been widely used in various fields of social sciences, and more and more researchers use it as a tool for empirical analysis. Compared with the previous statistical models, multilevel statistical model not only includes the information observed from the micro-level, but also contains the data obtained from a macro perspective. However, multilevel model does not work well, when it lacks data on each level. Therefore, the accumulated multilevel statistical model was established based on multilevel statistical analysis model and the gray system theory. At present, the cumulated multilevel statistical models can only solve a two-level structure with the small sample, but as for cross-classification two-level data and data with a three-level structure of the small sample, there is no established corresponding accumulated multilevel statistical models.Banana industry plays a significant role in China’s development. According to statistics, banana planting industry has provided over100million employment opportunities, and pushed the development of a series of relevant industries, such as the seed production, processing and transportation, product marketing and so on. How to better develop the banana industry has become one of the urgent problems to address. As a new form of organization and established in recent years, the banana industry data is scarce and very difficult to collect. The traditional statistical model cannot be used to analyze such kind of data, therefore this paper uses multi cross-classification accumulated multilevel statistical model to study Chinese banana industry.This paper collected a great number of data of banana industry of many areas from Hainan and Guangxi provinces and analyzed with cross-classification accumulated two-level statistical model and cross-classification accumulated three-level statistical model separately. The research results show that the type of interaction between regions and organizations has a significant effect on the yield of banana tissue, and regional per capita GDP, fertilization and control technology for banana organizational performance have a positive effect.