A Statistical Analysis of Subjective and Objective Methods of Evaluating Fabric Handle

A survey of the fabric handle preferences of panels of expert judges drawn from the textile and clothing industries of Japan, Australia, New Zealand and India has been undertaken. These data have been analysed using two statistical techniques, the first to select the most reliable judges and the second to analyse national judging characteristics.The technique for selecting the most reliable judges employs a linear model in which each judge's performance is tested against an orthogonal basis of fabric judgment calculated from the assessments of the remaining judges. In an iterative process, the judge whose hadle assessments are least explained by all the other expert judges is then sequentially eliminated until 8 of the original 53 judges remain. This final group was comprised of 4 judges from Japan, 3 from Australia and 1 from India. The results for the New Zealand judges indicate that they include some different characteristics when assessing fabric handle.The results from a principal component analysis of variance that although a single factor can account for a substantial amount of the observed variability in the data, additional factors are clearly implicated. The three most significant of these additional factors can be related to distinctive national handle preferences by the Japanese, New Zealand and Australian judging panels.