我国上市公司财务困境预警指标的检验及选取

Many papers blindly select warning indexes using the statistical methods which need financial indexes obey the test for normalized distribution, and that lots of literatures is not put up normal examinations for financial indexes, aiming at these defects ,this paper firstly uses K-S test、Q-Q probability chart putting up the test for normalized distribution of the 28 choosed financial indexes, it finds that the financial indexes is not accord with hypothesis of normalized distribution as a whole，accordingly，it ensures the followed-up method of financial indexes. Then this article tries to use non-parametric Wilcoxon test for mean value difference of the financial indexes, which is different from the method of T test mean difference of the financial indexes, accordingly，it is a certain degree of guidance to select the input variables. Finally, different from the other scholars using principal component analysis, stepwise discriminant analysis or Pearson correlation coefficient to choose financial indexes, this paper tries to use Spearman two-tailed t test for non-parametric correlation analysis for choosing the financial indexes of the smaller Spearman coefficients as the input variables. Warning indexes of the Listed Corporate Financial Distress selected by this method are more scientific and rational.