摘要
在排除模型误用以后,将经济学研究中针对同一问题的研究,不同计量模型的结果各不相同,甚至结论截然相反问题称为"计量模型多样性悖论"。分析此问题产生的原因,主要是关键变量的多样性、方程形式设定的多样性和计量模型的多样性,提出采用元分析来解决这个问题,指出在元分析应用过程中应力求单个模型的精确,尽可能多地采用合适的模型,注意同类性质的结果才能进行元分析。同时,讨论元分析带来的新问题,比如对研究团队的计量经济学水平要求较高,增加了成本,延长了研究的时间,也加大了论文的篇幅等,但这是计量经济学发展过程中的正常现象,并不涉及采用元分析解决"计量模型多样性悖论"问题的科学性。
This paper will be on the same problems in the research of economics study,the results of different econometric model each are not identical,even conclusion is in contrast to the problem is known as the paradox of econometric model diversity after the exclusion of econometric model misuse.It analyzes the main causes of this problem is the diversity of key variables,forms of equation and econometric model,and proposes to use meta analysis to solve this problem.It notes that we should seek a single accurate model,as much as possible using a suitable model,pay attention to the similar nature results in the metaanalysis application process.The authors also discuss the new problems caused by meta analysis,such as higher econometric level requirements for the research team,increased costs,extend the time to study,and also increased the length of the paper,etc.But the authors think this is the normal phenomenon in the process of econometrics development and it does not involve the use of meta analysis to solve the ＂econometric model diversity paradox＂ problem scientifically.
出处
《统计与信息论坛》
CSSCI
北大核心
2016年第11期20-26,共7页
Journal of Statistics and Information
基金
商务部重点课题<欠发达地区创新驱动发展的关键问题与对策研究>(2015SWBZD03)
关键词
计量经济
结论多样性
悖论
元分析
econometrics
diversity of conclusion
paradox
meta-analysis
