Types of statistical arguments in empirical research in psychology

Numerous solutions have been proposed to address the problem of statistical misuse and misunderstanding in research in psychology. This might have diverted attention from how or whether scientists are actually engaging in statistical reasoning by arguing about and with statistics within the context of the reasoning involved in empirical research in their written papers. In this paper, we studied statistical reasonings in psychological research papers. Specifically, we inductively coded arguments of 42 papers from seven highly cited psychology journals. With these data, we formed a partial, empirical typology with which we described different types of statistical arguments used in the literature. Papers, and their statistical arguments, in empirical psychology can be interpreted as possessing a 'legal' nature. Common arguments, such as statistical significance, are often used as a tacit and concise way of certifying that a claim, including conclusions, is empirically and statistically substantiated under the social standards of good research in our field. However, approval of research based on certifications may hinder the exposition and clarity of statistical arguments.