Interval Econometrics And Applications For Panel Data

Interval-value data has attracted the attention of scholars at home and abroad since it was proposed and has widely used in climate,medicine,finance and other fields.Because interval-value data is to observed and recorded the rang of values rather than a single point value data,it is helpful to understand the deviation and variation of data,and can also be used to describe the uncertainty of variables.According to the literature,most of the modeling methods of interval-value data are combined with the modeling methods of cross-section data or time series data,while few of the research results are combined with the panel data modeling methods.In this paper,considering that the panel data is two-dimensional data,it contains cross-section data information and time series data information,which can be easily obtained more data samples and collinearity between degrees of freedom and reduce variables.Therefore,we set up the interval-value panel data model combined with the modeling methods of interval-value data and panel data.It provides a new perspective for analyzing panel data.Firstly,we generalize the following four linear regression models and give the detailed derivation process for interval-value data: Center Method(CM),Min Max Method(Min Max),Center and Range Method(CRM)and Paremetrized Method(PM).These methods have been used to discuss interval-value cross-section data.But we generalize these methods and apply them to analysis of interval-value panel data.What's more,we give the relationship between the four methods and display the results predicted by PM.Secondly,Monte Carlo simulation experiment is carried out to verify the effectiveness of the four methods.We set up five evaluation indicators and show the results in figures.It can be seen that the prediction results of PM is better than CM,Min Max and CRM,and support the theoretical analysis.Finally,an empirical example applying all four generalized methods is considered.We analyze and study the comprehensive effects of climate change and agricultural production factors on grain yield in China from 1993 to 2018.Some scientifically sound basis and suggestions are provided for China's grain stability and safe production.Moreover,the regression results of the four methods are given and compared,and the conclusion that PM is more suitable to describe this dataset is further verified empirically.