Criticism of Null Hypothesis Significance Testing: Limitations and Possible Ways Out. Part I

Abstract

The most common mistakes in using of Null-hypothesis significance testing (NHST) as a mean of data analysis in psychology are disclosed in the article. Several levels of criticism of NHST in recent publications are described: the level of presuppositions and main procedures of the approach; the level of social implication and after-effect; the level of psychological interpretation of statistical results. The main alternative proposals are analyzed. The preliminary conclusion is that not procedures of single study evaluation have to be changed but new structure of evaluation and comparison different researches results must be developed in the society.