期刊
JOURNAL OF EXPERIMENTAL PSYCHOLOGY-GENERAL
卷 151, 期 1, 页码 213-223出版社
AMER PSYCHOLOGICAL ASSOC
DOI: 10.1037/xge0001069
关键词
outliers; false-positive; methodology; statistical analysis; boxplot
Researchers should be cautious when choosing to exclude outliers, as condition-level exclusions may lead to inflated false-positive rates. The study highlights the importance of blind outlier exclusion procedures to avoid Type I errors.
When researchers choose to identify and exclude outliers from their data, should they do so across all the data, or within experimental conditions? A survey of recent papers published in the Journal of Experimental Psychology: General shows that both methods are widely used, and common data visualization techniques suggest that outliers should be excluded at the condition-level. However, I highlight in the present paper that removing outliers by condition runs against the logic of hypothesis testing, and that this practice leads to unacceptable increases in false-positive rates. I demonstrate that this conclusion holds true across a variety of statistical tests, exclusion criterion and cutoffs, sample sizes, and data types, and shows in simulated experiments and in a reanalysis of existing data that by-condition exclusions can result in false-positive rates as high as 43%. I finally demonstrate that by-condition exclusions are a specific case of a more general issue: Any outlier exclusion procedure that is not blind to the hypothesis that researchers want to test may result in inflated Type I errors. I conclude by offering best practices and recommendations for excluding outliers.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据