Comparative Statistics in Endodontics: A Narrative Review of t-test and Analysis of Variance

Statistical analysis is central to evidence-based dental and medical research. In endodontics, where studies often compare outcomes such as postoperative pain, anesthetic efficacy, or healing rates, appropriate test selection is essential for drawing valid conclusions. Misapplication of statistical tools can lead to misleading interpretations and diminish the impact of otherwise well-designed studies. This narrative review aims to simplify the concepts of the t-test and analysis of variance (ANOVA) for postgraduate students in dentistry, with a focus on practical application in endodontic research. The t-test is suitable for comparing the means of two groups, with independent samples t-test applied to separate groups (e.g., pain scores in male vs. female patients) and paired t-test used for repeated measures in the same group (e.g., pre vs. postinjection pain levels). Worked examples illustrate stepwise calculation and interpretation. ANOVA extends this comparison to three or more groups, preventing inflated error rates from multiple t-tests. The types include one-way ANOVA, repeated-measures ANOVA, and two-way ANOVA, each explained through endodontic clinical scenarios. When assumptions of normality and homogeneity of variance are violated, nonparametric alternatives – Mann–Whitney U, Wilcoxon signed-rank, Kruskal–Wallis, and Friedman tests – are recommended. Correct selection and interpretation of statistical tests enhance the quality and reliability of endodontic research. Clear reporting, including P values, effect sizes, and confidence intervals, ensures clinical relevance. By grounding statistical principles in real clinical examples, this review provides postgraduate students with a practical framework to apply t-tests and ANOVA confidently, thereby strengthening their contribution to evidence-based endodontics.