Approximate confidence intervals for effect sizes
Approximate confidence intervals for effect sizes
复制标题
DOI：
10.1177/0013164403256358
复制
发表时间：
2003-08-01
影响因子：
2.7
通讯作者：
Keselman, HJ
中科院分区：
文献类型：
--
作者：
Algina, J;Keselman, HJ
本文定义了相关（重复测量）组设计中效应大小的近似置信区间。作者发现他们的方法比伯德提出并承认的近似区间准确得多。也就是说，所有研究条件下的覆盖概率非常接近理论值 0.95。相比之下，伯德区间的覆盖概率远低于 0.95。此外，作者的区间不太可能像伯德的方法那样对效果做出过于乐观的描述。他们还检查了独立群体设计中效应大小的伯德区间的操作特征，发现虽然其覆盖概率的近似相当准确，但近似的准确性确实随着总体效应大小的大小而变化。
This article defines an approximate confidence interval for effect size in correlated (repeated measures) groups designs. The authors found that their method was much more accurate than the interval presented and acknowledged to be approximate by Bird. That is, the coverage probability over all the conditions investigated was very close to the theoretical .95 value. By contrast, Bird's interval could have coverage probability that was substantially below .95. In addition, the authors' interval was less likely than Bird's method to present an overly optimistic portrayal of the effect. They also examined the operating characteristics of the Bird interval for effect size in an independent groups design and found that, although it is fairly accurate in its approximation of coverage probability, the accuracy of the approximation does vary with the magnitude of the population effect size.