作者: Joel R. Levin
DOI:
关键词: Data dredging 、 Educational research 、 Process (engineering) 、 Function (engineering) 、 Computer science 、 Statistical hypothesis testing 、 Econometrics 、 Statistical significance 、 Empirical research 、 Risk analysis (engineering) 、 Test use
摘要: Questions and concerns are directed to those who advocate replacing statistical hypothesis testing with alternative data-analysis strategies. It is further suggested that: (1) commonly recommended hypothesis-testing alternatives anything but perfect, especially when allowed stand alone without an accompanying inferential filtering device; (2) various modifications can be implemented make the process its associated conclusions more credible; (3) testing, intelligently, adds importantly story-telling function of a published empirical research investigation.