Department of Crop Sciences, University of Illinois, Urbana, IL, USA.
Pest Manag Sci. 2022 Mar;78(3):847-854. doi: 10.1002/ps.6668. Epub 2021 Oct 22.
P values, binary hypothesis tests, and statistical significance are too often overused or used incorrectly in pest management reports. These statistical results are not the point of an analysis, they are just a by-product that can sometimes be informative. The point of the analysis is always a biological interpretation of the data focusing on whether something new has been learned and then quantifying what was learned in terms of biological effects or supported hypotheses. The term 'statistical significance' should be retired if only because of the common misunderstanding that it implies biological relevance; these terms are not equivalent. © 2021 Society of Chemical Industry.
在害虫管理报告中,P 值、双假设检验和统计显著性经常被过度使用或使用不当。这些统计结果不是分析的重点,它们只是一个副产品,有时可能具有信息性。分析的重点始终是对数据的生物学解释,重点是是否有新的发现,然后根据生物学效应或支持的假设来量化所学到的内容。如果仅仅因为普遍存在对“统计显著性”的误解,即它暗示着生物学相关性,那么“统计显著性”这个术语应该被废弃;这些术语并不等同。 © 2021 化学工业学会。