Prediction of fabric handle value using ordinal regression model

Abstract

Bed sheet fabric as a kind of home textile has been used since many years ago. Bed sheet is very significant because of being in direct contact with body consecutively for a long period of time. Bed sheet surplus qualitative parameters such as fiber substance, method of printing, finishing, etc., have a significant parameter called handle. In this paper, we proceeded to consider the relationship between fabric handle as a qualitative parameter and physical parameters which influenced the fabric handle using statistical modeling. The statistical model used was ordinal regression model. The modeling was done by SPSS V.19 software. We used 15 bed sheet fabrics. For subjective evaluation of 15 bed sheet fabrics, we selected 55 persons randomly as sample members according to Cochran’s formula. Population was selected from senior BS students and MS students at Isfahan University of Technology (IUT). We asked persons to classify bed sheet fabrics based on their preference of fabric handle from 1 (lowest) to 5 (highest). Physical parameters values were obtained through standard experiments. Finally, we analyzed obtained data through SPSS V.19 using ordinal regression model. Results showed a satisfying match between extracted data from the software and the real data from person’s evaluation.

Keywords:

• bed sheet fabric
• subjective evaluation
• handle value
• ordinal regression model
• SPSS