Chapter 56: Evaluating Hypothesis Tests: Test Size and Power
作者
Mary C. Meyer
出处
期刊:Society for Industrial and Applied Mathematics eBooks [Society for Industrial and Applied Mathematics] 日期:2019-01-01卷期号:: 481-488
标识
DOI:10.1137/1.9781611975789.ch56
摘要
When planning and designing statistical studies, researchers are interested in the power of tests. Once the appropriate statistical model is determined, the power of a test depends on a number of factors, including the test statistic, test size, sample size, and the “effect size,” or how different the truth is from the null hypothesis. Other parameters that might be unspecified by the hypotheses, such as the variance when testing the mean of a normal population, can also affect the power. After choosing a test statistic and a desired size α, researchers want to know ahead of time what the power would be for various guesses at effect sizes and values for other parameters, for a range of sample sizes, in order to choose a sample size that is likely to have adequate power.