Crone Gabriel, Green Christopher D
York University.
Theory Psychol. 2025 Jun;35(3):359-380. doi: 10.1177/09593543241311861. Epub 2025 Feb 1.
In psychology, it is largely assumed that researchers collect real data and analyze them honestly-that is, it is assumed that data fabrication seldom occurs. While data fabrication is a rare phenomenon, estimates suggest that it occurs frequently enough to be a concern. To this end, statistical tools have been created to detect and deter data fabrication. Often, these tools either assess raw data, or assess summary statistical information. However, very few studies have attempted to review these tools, and of those that have, certain tools were excluded. The purpose of the present study was to review a collection of existing statistical tools to detect data fabrication, assess their strengths and limitations, and consider their place in psychological practice. The major strengths of the tools included their comprehensiveness and rigor, while their limitations were in their stringent criteria to run and in that they were impractical to implement.
在心理学领域，人们普遍认为研究人员收集的是真实数据并诚实地进行分析，也就是说，人们认为数据造假很少发生。虽然数据造假是一种罕见的现象，但据估计其发生频率足以令人担忧。为此，已经创建了统计工具来检测和阻止数据造假。通常，这些工具要么评估原始数据，要么评估汇总统计信息。然而，很少有研究试图对这些工具进行综述，而在那些进行过综述的研究中，某些工具被排除在外。本研究的目的是对现有的一系列检测数据造假的统计工具进行综述，评估它们的优缺点，并考虑它们在心理学实践中的地位。这些工具的主要优点包括其全面性和严谨性，而它们的局限性在于运行标准严格，且实施起来不切实际。
