Statistical reporting practices within forensic psychology: Is anything changing?

Abstract

Objective: We examined the evolution of statistical reporting practices within forensic psychology across two decades (2000-2020) to assess their adherence to recommended best practices.

Method: We conducted a comprehensive analysis of articles published in six prominent forensic psychology journals, including every empirical article published in each journal in 2000, 2005, 2010, 2015, and 2020 (N = 813). We then evaluated the use and interpretation of null hypothesis significance testing (NHST), effect sizes (ESs), confidence intervals (CIs), and Bayesian statistics for each article in the sample.

Results: We found a persistent reliance on NHST, with nearly all articles employing it for data analysis and interpretation. Encouragingly, the reporting of ESs and CIs has increased substantially; their interpretative use, however, remains limited. Bayesian methods were rarely used for analysis or interpretation of data.

Conclusions: These findings suggest a slow uptake of reforms advocated by statistical guidelines, with forensic psychology researchers continuing to prioritize NHST over recommended approaches. Although the increase in ES and CI reporting is encouraging, the continued reliance on NHST means that both the scientific literature and important applied decision making in the forensic psychology field are impacted by the shortcomings of this statistical reporting approach (e.g., simplistic dichotomous decisions, lack of reproducibility, potential for p-hacking). We call for journals in the field to encourage further use of statistical best practices within their manuscripts. (PsycInfo Database Record (c) 2025 APA, all rights reserved).