统计研究

• 论文 • 上一篇    下一篇

横截面数据变量的规模特征:特征价格建模分析

秦朵 刘一萌   

  • 出版日期:2015-02-15 发布日期:2015-03-17

Modelling Scale Effect in Cross-section Data: The Case of Hedonic Price Regression

Qin Duo & Liu Yimeng   

  • Online:2015-02-15 Published:2015-03-17

摘要: 本文通过对截面数据排序这样一种创新的简单建模实验,将经济变量间关系的一个基本特征——非线性规模效应纳入到截面数据模型设定中。我们选取特征价格模型为实验案例,具体的分析对象是汽车和电脑的特征价格模型。实验得出的主要结论有:(1)由于忽略了截面数据样本中潜在的非线性规模效应,传统模型得出的特征价格指数很可能存在系统偏差;(2)基于规模解释变量的数据排序方法是滤出截面数据样本中非线性信息的一种简易而有效的途径;(3)截面数据一经排序,便可采用现有的系统动态建模方法来实现对变量间这种非线性规模信息的滤出。

关键词: 截面数据, 数据排序, 变量规模特征, 特征价格, 共因子模型

Abstract: An innovative and simple modelling experiment with cross-section data ordering is carried out to exploit a basic feature among many economic variables – nonlinear scale effect and incorporate the effect into specifications of cross-data models. The experiment is conducted on hedonic price models using two data sets of automobiles and computers respectively. The key findings are: (1) Hedonic price indices can be significantly biased if they are constructed using models which disregard possible nonlinear scale effect underlying in data samples; (2) Scale-based data ordering offers considerable potential to filter such scale-dependent nonlinear information from cross-section samples; (3) The filtering can be easily carried out by systematic adoption of dynamic modelling methods once the cross-section data is ordered.

Key words: Cross-section Data, Data Ordering, Scale Effect, Hedonic Price, COMFAC Model