摘要
假设检验这一传统的处理实验数据的方法,目前在心理学中正经受着普遍的批评。在这一背景下,一些研究者提出一些改进假设检验的方法,其中最典型的是Harris提出的假设检验方法;另一些研究者则提出用其他方法取代假设检验,其中尤为人们所强调的是效应值估计。在分析这两种取向的基础上,该文章指出最好的报告心理学实验结果的方法是置信区间,因为它能够把统计推断和效应估计整合在一起。
Hypothesis testing,as a typical method of reporting the results of psychological experiments,is now receiving popular criticism in psychology.As a result of this,some researchers offer several revised methods of the classical hypothesis testing,the typical one of which is Harris＇ hypothesis testing paradigm;some others suggest that the hypothesis testing should be replaced by other data-analysis methods,one of which that is mostly emphasized is effect-size estimation.Based on the analysis of the two approaches,it is suggested that the best way to report the results of psychological experiments should be confidence interval,which can integrate statistic inference and effect-size estimation.
出处
《心理学探新》
CSSCI
2010年第5期62-65,共4页
Psychological Exploration
关键词
假设检验
效应估计
置信区间
hypothesis testing
effect-size estimation
confidence interval
