Large-Scale Hypothesis Testing and Confidence Interval Estimation: An Effect-Size and Multiplicity-Aware Framework
Qubahan Journal of Coaching and Sports Sciences · 15.10.2026 · 4 (4) · 1-16
DOI: 10.48161/qjcs.v4n4a90 ↗Kurz gefasst
Fachartikel in der Zeitschrift Qubahan Journal of Coaching and Sports Sciences, online erschienen am 15.10.2026. Verfasst von Ibrahim Mallam Fali und 1 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
This study develops and empirically evaluates an integrated framework for statistical hypothesis testing and confidence interval estimation in large-scale data analysis. Using the 2015 Behavioral Risk Factor Surveillance System (BRFSS) binary diabetes health-indicator file, the analysis examined 253,680 observations and 22 variables, including diabetes status, cardiometabolic risk indicators, lifestyle indicators, healthcare-access measures, and socioeconomic variables. The analysis used descriptive statistics, prevalence estimates, Welch mean-comparison tests, chi-square tests, confidence intervals, standardized effect sizes, odds ratios, risk differences, multiple-comparison correction, and adjusted logistic regression. Diabetes prevalence in the main file was 13.93% (95% CI, 13.80% to 14.07%). The results showed that the large sample size produced very small p-values for most comparisons; however, effect magnitudes varied substantially across variables. Body mass index (BMI) was higher among respondents with diabetes by 4.14 units, with a 95% confidence interval of 4.06 to 4.22 and a moderate standardized effect. High blood pressure showed a strong unadjusted association with diabetes (odds ratio 5.04; 95% CI, 4.91 to 5.17), while physical activity showed a lower odds pattern for diabetes. After adjustment, high blood pressure, high cholesterol, general health, age, and BMI remained important predictors. The study demonstrates that hypothesis testing is most reliable when p-values are interpreted together with confidence intervals, effect sizes, adjusted estimates, and multiplicity-aware decision rules. The framework provides a reproducible model to strengthen statistical interpretation in large empirical datasets.
Zitieren
APA
Fali, I. M., & Nnenna, E. C. (2026). Large-Scale Hypothesis Testing and Confidence Interval Estimation: An Effect-Size and Multiplicity-Aware Framework. Qubahan Journal of Coaching and Sports Sciences, 4(4), 1-16. https://doi.org/10.48161/qjcs.v4n4a90
Vancouver
Fali IM, Nnenna EC. Large-Scale Hypothesis Testing and Confidence Interval Estimation: An Effect-Size and Multiplicity-Aware Framework. Qubahan Journal of Coaching and Sports Sciences. 2026;4(4):1-16. doi:10.48161/qjcs.v4n4a90
BibTeX
@article{Fali2026,
author = {Fali, Ibrahim Mallam and Nnenna, Eze Claris},
title = {Large-Scale Hypothesis Testing and Confidence Interval Estimation: An Effect-Size and Multiplicity-Aware Framework},
journal = {Qubahan Journal of Coaching and Sports Sciences},
year = {2026},
volume = {4},
number = {4},
pages = {1--16},
doi = {10.48161/qjcs.v4n4a90}
}
EUVIMED