Moving Beyond the Null Hypothesis Significance Testing Controversy: Guidance for Doctoral Research Supervisors
, Susan K Steele-MosesThe aim of this article is to provide doctoral research supervisors with a practical pedagogical toolkit for guiding students toward modern estimation thinking and precision-based sample planning. The toolkit is intended to reduce over-reliance on null hypothesis significance testing (NHST) and to enhance dissertation interpretability, reproducibility, and meta-analytic value.
Despite decades of NHST criticism, doctoral programs often require quantitative dissertation research within an NHST framework. An estimation approach that focuses on the point estimate and confidence interval is philosophically different from the dichotomous thinking in NHST. Doctoral research supervisors face challenges when guiding their students through this controversy, as many students understand statistics from an NHST perspective. This knowledge gap creates challenges when planning dissertation research and obtaining committee approval.
Through a synthesis of seminal articles, statistics texts, and recent publications, a pedagogical resource for research supervisors is presented that explains difficult concepts such as p value, conditional probability, alpha (α), beta (β), point estimate, confidence intervals, and precision-based sample planning. Key concepts are illustrated using relatable examples such as a pain scale and the Beck Depression Inventory.
This pedagogical resource presents a supervisory toolkit that clarifies the distinction between traditional NHST approaches and contemporary estimation thinking and supports doctoral research supervisors in planning, interpreting, and reporting quantitative studies. The paper includes three core pedagogical tools: (1) “How-To” boxes to support supervisory practice, (2) applied examples using familiar measures (pain scales and the BDI-II), and (3) a staged roadmap of key supervisor questions across the dissertation process. It also provides four technical resources: (4) Appendix A (mapping effect-size indices to 10 common statistical tests), (5) Appendix B (manual computation formulas with interpretive benchmarks), (6) Appendix C (NHST quiz answers and explanations), and (7) precision-based planning checklists.
The synthesis demonstrates that applying estimation thinking enhances interpretation, promotes precision-based sampling, and improves the reporting and pedagogical coherence of quantitative dissertation research.
Research supervisors should help novice researchers understand and apply estimation approaches when supervising dissertation research. When an NHST framework is primarily used, supplementing it with an estimation approach will aid in planning, execution, and reporting quantitative results. This integrated approach helps students recognize the value of their research beyond dichotomous NHST interpretations.
Recommendations consistent with APA guidelines assist research supervisors in guiding their students by evaluating design options, planning research, understanding and reporting statistical results, and aligning with cumulative meta-analytic literature.
When research supervisors guide their students to apply estimation thinking, the resulting research is more interpretable, reproducible, and actionable. Estimation results support practical significance by providing effect sizes and confidence intervals that can be meaningfully interpreted. By ensuring these values can also be aggregated across studies, estimation thinking enhances the meta‑analytic value of dissertation research. Understanding the distinction between statistical significance and practical significance strengthens the societal impact of research findings, enhancing dissertation reproducibility, interpretability, and meta-analytic value for cumulative science.
Future research should examine how statistical cognition in estimation thinking develops among doctoral researchers under different supervisory approaches and continue to refine pedagogical resources that promote understanding of estimation thinking.
Back