摘要
目的探讨单因素重复测量设计与随机区组设计资料的方差分析的关系。方法利用文献数据[1,2],拟对资料进行单因素重复测量设计和随机区组设计2种处理,比较2种处理结果的异同。结果单因素重复测量设计资料在满足球对称的条件和正态对称条件下,与误用随机区组设计处理结果相同;尤其在球不对称且F值F界值时,采用随机区组设计处理与单因素重复测量分析校正前后的结果,也会出现一致情况。结论2种方法都是单因素试验的方差分析,但2种设计的方法和分析目的均不同,不可混同。
Objective To explore the relation of one-way repeated measures design ANOVA and randomized block design ANOVA. Methods To analyze data quoted from documents by one-way repeated measures design and randomized block design respectively and compare the differences of the two results. Results Under the conditions of data satisfied the normal symmetry and sphericity symmetry, to get the same results by one-way repeated measures design method and randomized block design misused, especially under the condition of sphericity dissymmetry and F value 〉〉F criteria, the same results were calculated too. Conclusion The two types of statistic methods are all one-way ANOVA, but their analyzed objectives are different and can not be misused.
出处
《中国医院统计》
2008年第2期102-104,共3页
Chinese Journal of Hospital Statistics
关键词
重复测量
随机区组设计方差分析
Repeated measures design Randomized block design ANOVA