摘要
本文运用四川省28个县医院的资料对医院评价指标进行了筛选方法探讨。在使用传统筛选指标方法的同时,又提出了一种新的代表性指标筛选方法——类成分法,并比较了各种方法的优缺点,对指标的代表性进行了尝试性验证,结果表明类成分法较合理,由该法选出的指标既保证了类内指标的代表性又保证了类间指标的独立性。
In this article, based on the data of 28 county hospitals, we applied traditional method of correlation coefficient, clustering analysis and principal components to select hospital-evaluation indexes and analysed the merits and demerits of these methods, meanwhile putting forward a new method-clustering Component analysis to choose typical indexes, comparing it with the above methods and verifying the typical indexes. The analysis indicated that the typical indexes selected by the new method possessed most information of the original index, and were independent of one another , corresponding to the reality.
出处
《中国医院统计》
1995年第4期193-196,共4页
Chinese Journal of Hospital Statistics