COMPARATIVE EVALUATION OF DATA TRANSFORMATION TECHNIQUES FOR NORMALIZING EARLY-AGE CONCRETE COMPRESSIVE STRENGTH DATA
International Journal of Scientific Engineering and Technological Research and Innovations · 01.01.2025
DOI: 10.66837/ijsetri01/489 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Scientific Engineering and Technological Research and Innovations, online erschienen am 01.01.2025.
Abstract (Originalsprache)
The assumption of normality underpins many statistical procedures in construction material analysis, yet empirical datasets frequently violate this condition. This study evaluated the effectiveness of three transformation techniques—Manly, Yeo–Johnson, and Nelson—in normalizing 7-day concrete compressive strength data. Normality was assessed using Shapiro–Wilk and Anderson–Darling tests, alongside skewness and kurtosis statistics. The raw dataset exhibited extreme non-normality (W = 0.1329, p < 0.001; A = 44.158, p < 0.001), characterized by severe positive skewness and heavy tails. Among the transformations, the Manly method achieved near-perfect normalization (Shapiro–Wilk p = 0.9726), with skewness and kurtosis approximating theoretical normal values. In contrast, Yeo–Johnson and Nelson transformations failed to sufficiently correct distributional distortions. The findings highlight the importance of empirical validation in transformation selection and establish the Manly transformation as the most effective technique for this dataset.
Zitieren
APA
(2025). COMPARATIVE EVALUATION OF DATA TRANSFORMATION TECHNIQUES FOR NORMALIZING EARLY-AGE CONCRETE COMPRESSIVE STRENGTH DATA. International Journal of Scientific Engineering and Technological Research and Innovations. https://doi.org/10.66837/ijsetri01/489
Vancouver
COMPARATIVE EVALUATION OF DATA TRANSFORMATION TECHNIQUES FOR NORMALIZING EARLY-AGE CONCRETE COMPRESSIVE STRENGTH DATA. International Journal of Scientific Engineering and Technological Research and Innovations. 2025. doi:10.66837/ijsetri01/489
BibTeX
@article{euvimed2025,
title = {COMPARATIVE EVALUATION OF DATA TRANSFORMATION TECHNIQUES FOR NORMALIZING EARLY-AGE CONCRETE COMPRESSIVE STRENGTH DATA},
journal = {International Journal of Scientific Engineering and Technological Research and Innovations},
year = {2025},
doi = {10.66837/ijsetri01/489}
}
EUVIMED