University of Manitoba.
J Appl Behav Anal. 1974 Winter;7(4):643-5. doi: 10.1901/jaba.1974.7-643.
The thrust of this paper is to bring to the attention of operant researchers statistical procedures that are appropriate for correlated data. In addition to specifying these statistical procedures consideration is given to the question of using individual comparison statistics rather than omnibus F tests. Specifically, it is recommended that a more powerful test of the experimental hypotheses can be obtained by performing Bonferroni t statistics rather than analysis of variance F tests.
本文的主旨是向操作性研究人员介绍适用于相关数据的统计程序。除了指定这些统计程序外,还考虑了使用个体比较统计而不是总 F 检验的问题。具体来说,建议通过进行 Bonferroni t 检验而不是方差分析 F 检验来获得更有力的实验假设检验。