Differences Observed in Data Analysis Techniques: An Example Using Natural Fibers’ Diameters and Absorption Times

Abstract

<abstract> <bold><sc>Abstract. </sc></bold>The properties of natural fibers are often evaluated across species, with the average value of the property of interest being compared via analysis of variance (ANOVA). Researchers also frequently adjust property values to account for the natural variation existing between some characteristic of the fibers. Weighting the property to correct for the variation, then using ANOVA, or applying analysis of covariance (ANCOVA) are two alternative data analysis methods. While studying the surface characteristics of bamboo, jute, ramie, and hemp fibers, observations were made on the differences in absorption time of the natural fibers. These differences were then compared using three data analysis methods- ANOVA using the average and an adjusted average based on the diameter of the fiber for multiple comparisons, and an adjusted average following an analysis of covariance using fiber diameter as the covariate. Only hemp’s mean absorption time remained in the same t grouping regardless of the method employed. The other species’ average absorption times changed t groupings depending on the method chosen. This simplified example highlights the value of the method chosen for data analysis.