Advancing Statistical Literacy in Eye Care: A Series for Enhanced Clinical Decision-Making (Biostatistik in Ophthalmologie und Optometrie: Eine Artikelserie zur Unterstützung evidenzbasierter Entscheidungsprozesse): Part 2: Statistical Testing in Eye and Vision Research – Principles and Practices (Teil 2: Statistische Tests – Prinzipien und Anwendung)

Purpose: To enhance statistical literacy in eye and vision Research through practical guidance on hypothesis testing, test selection, and interpreting inferential statistics. The goal is to reduce common analytical errors and support more clinically meaningful conclusions in ophthalmic studies.

Material and Methods: A narrative literature review was conducted using PubMed, Scopus, and Web of Science to identify best practices in hypothesis testing, error control, and test selection relevant to clinical research in ophthalmology
and optometry. Simulated datasets, based on real-world clinical scenarios, were generated in Python to illustrate core concepts. Worked examples demonstrate the Impact of sample size, data distribution, and error type on statistical
conclusions.

Results: Common misinterpretations of p-values and frequent misuse of statistical tests were identified. The review explains how reducing the significance level (α) increases Type II error risk unless the sample size is increased. A structured decision framework was developed to aid the choice between parametric and non-parametric tests, including when assumptions are violated. Simulations and clinical examples demonstrate how effect size, variability, and multiple testing adjustments affect results.

Conclusion: Statistical missteps in eye research often arise from poor test selection, inadequate power, or overreliance on p-values without context. This article advocates for the use of confidence intervals, effect sizes, and ransparent
reporting to enhance the credibility of research findings. Following structured analytic frameworks and established reporting guidelines (e. g. CONSORT, STROBE) helps ensure that statistical conclusions align with clinical relevance, ultimately supporting better patient care and more trustworthy research outcomes.