(广东商学院国民经济研究中心，广东商学院华商学院)

【摘要】传统的因子分析模型，公因子解不能降维或失去一些变量解释，故因子分析需要找出一个优良模型。而优良模型的确定需要更好的量化优良性准则，为此本文用因子分析的基本思想和目的，提出因子分析降维性、解释性、拟合性更好的量化优良性准则，并证明有目标函数最大化的因子分析模型L是因子分析的一个优良模型和方法。实证表明，因子分析模型L具有优良性，传统因子分析模型的最大似然法、主因子法误差较大。

关键词  因子分析模型L  优良模型  实证分析

中图分类号  O212       文献标识码  A

The Superiority and Application of

The Factor Analysis Model L

Abstract: Because of the prevalence of the factor analysis model, common factor solution not being able to reduce dimensionality or an explanation of the loss of a number of variables, therefore, the factor analysis needs to find a better model, and the confirmation of a better model  needs for a good standard. To this end, here it is used the basic idea of the factor analysis to put forward its dimensionality reduction, explanation, and the superior standards of fit quantitative evidence. It proves that the Factor Analysis Model L that has maximize the objective function is a better model for factor analysis.The actual data has been illustrated: Factor Analysis Model L is a better method, the maximum likelihood method and the principal factor method is sometimes of a bigger error.

Key words: Factor Analysis Model L；Super Model；Empirical Analysis