Document Type : Original Article
Authors
1
Department of Sports Science, Faculty of Humanities, Pa.C., Islamic Azad University, Parand, Tehran, Iran
2
Department of Statistics, Faculty of Mathematical and Computer Sciences, Shahid Chamran University, Ahvaz, Khuzestan, Iran
10.22034/spmi.2026.572261.2838
Abstract
Distinguishing between statistical significance and practical importance remains a persistent challenge in interpreting findings in exercise physiology research. An overreliance on p-values may promote dichotomous interpretations, obscure statistical uncertainty, and potentially lead to suboptimal training or clinical decisions. The purpose of this meta-research study was to examine statistical reporting practices and to evaluate the correspondence between statistical significance and practical importance in Iranian exercise physiology articles. Using a systematic meta-research design, full-text articles published between 2019 and 2025 in Iranian exercise physiology journals were reviewed. After screening, 45 articles were included in the final analysis. Each article was coded for the reporting of p-values, exact p-values, effect sizes, uncertainty measures, and the availability of sufficient numerical data for effect size reconstruction. When possible, standardized mean differences corrected for small-sample bias were calculated along with their 95% confidence intervals, and results were categorized using a 2×2 framework (statistically significant vs. non-significant × practically important vs. not important). Most articles reported p-values, but rarely included effect sizes and confidence intervals. Estimation-based reanalyzes showed a limited relationship between statistical significance and practical importance: some statistically significant results had trivial effects, while some non-significant results had potentially meaningful effects. These findings show the limits of relying on thresholds and the risk of confusing statistical significance with real impact. In conclusion, the results highlight the need to shift from p-value–centered reporting to an estimation-focused approach that emphasizes effect sizes and uncertainty. Such a shift is particularly important in exercise physiology, where applied decisions depend on an accurate understanding of both the magnitude and the precision of observed effects.
Keywords