Reevaluating p-value and its Impact on Conventional Results Display and Interpretation

Abstract

There are numerous instances of misunderstanding, misinterpreting, and misusing p-values, which have raised concerns about the reliability of statistical conclusions based on them. This paper provides a comprehensive reevaluation of the mathematical foundations of the p-value, discusses common misconceptions in its teaching and application, and explores alternatives such as confidence intervals, effect size, and Bayes’ factor. Additionally, it highlights scenarios where the p-value remains valuable beyond hypothesis testing. The goal is to emphasize the continued relevance of the p-value while acknowledging potential pitfalls in its misuse. Given that most of the discussion is rooted in applied statistics, this paper aims to offer clarity to researchers across various disciplines.