当前位置: X-MOL 学术Social and Personality Psychology Compass › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Have psychologists increased reporting of outliers in response to the reproducibility crisis?
Social and Personality Psychology Compass ( IF 4.8 ) Pub Date : 2021-03-24 , DOI: 10.1111/spc3.12591
Kathrene D. Valentine 1, 2 , Erin M. Buchanan 3 , Arielle Cunningham 4 , Tabetha Hopke 4 , Addie Wikowsky 4 , Haley Wilson 4
Affiliation  

Psychology is currently experiencing a “renaissance” where the replication and reproducibility of published reports are at the forefront of conversations in the field. While researchers have worked to discuss possible problems and solutions, work has yet to uncover how this new culture may have altered reporting practices in the social sciences. As outliers and other errant data points can bias both descriptive and inferential statistics, the search for these data points is essential to any analysis using these parameters. We quantified the rates of reporting of outliers and other data within psychology at two time points: 2012 when the replication crisis was born, and 2017, after the publication of reports concerning replication, questionable research practices, and transparency. A total of 2235 experiments were identified and analyzed, finding an increase in reporting from only 15.7% of experiments in 2012 to 25.0% in 2017. We investigated differences across years given the psychological field or statistical analysis that experiment employed. Further, we inspected whether data exclusions mentioned were whole participant observations or data points, and what reasons authors gave for stating the observation was deviant. We conclude that while report rates are improving overall, there is still room for improvement in the reporting practices of psychological scientists which can only aid in strengthening our science.

中文翻译:

为了重现可再生性危机,心理学家是否增加了离群值的报告?

心理学目前正在经历一种“复兴”,其中已发表报告的复制和可再现性处于该领域对话的最前沿。尽管研究人员一直在讨论可能的问题和解决方案,但尚未发现这种新文化如何改变社会科学报告实践的工作。由于异常值和其他错误数据点可能会使描述性统计数据和推断性统计数据产生偏差,因此对于使用这些参数进行的任何分析,搜索这些数据点都是必不可少的。我们在两个时间点对心理学上离群值和其他数据的报告率进行了量化:2012年是复制危机诞生的时期,而2017年是在发布有关复制,可疑的研究实践和透明度的报告之后。总共鉴定和分析了2235个实验,发现报告的数量从2012年的15.7%增加到2017年的25.0%。鉴于实验所采用的心理领域或统计分析,我们调查了多年来的差异。此外,我们检查了提到的数据排除是整个参与者的观察结果还是数据点,以及作者给出陈述观察结果的原因是与众不同的。我们得出的结论是,虽然报告率总体上在提高,但心理学家的报告实践仍有改进的空间,这只能有助于加强我们的科学。我们检查了提到的数据排除是整个参与者的观察结果还是数据点,以及作者给出陈述观察结果的原因是与众不同的。我们得出的结论是,虽然报告率总体上在提高,但心理学家的报告实践仍有改进的空间,这只能有助于加强我们的科学。我们检查了提到的数据排除是整个参与者的观察结果还是数据点,以及作者给出陈述观察结果的原因是与众不同的。我们得出的结论是,虽然报告率总体上在提高,但心理学家的报告实践仍有改进的空间,这只能有助于加强我们的科学。
更新日期:2021-05-15
down
wechat
bug