Study on a Quantified Approache to the Hand-Felt Softness of Facial Tissue - the Application of Reciprocal Matrix Method

• Main Authors: Teng, TserYing, 鄧澤殷
• Other Authors: Perng, YuanShing
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/04478508731731608493

博士 === 大葉大學 === 環境工程學系碩士班 === 99 === In previous studies of household tissue softness using the hand-felt approach, n2 or n - 1 paired comparisons and ranking methods were employed to distinguish the order of the samples; however, subjective perceptions and objective sample deviations could not easily be excluded. In the present study, two research purposes were set in order to improve the drawbacks: 1) Establish SOP simulated regular using of end users to test hand-felt softness of facial tissue. 2) Applicate reciprocal matrix to quantify and infer hand-felt softness. The SOP inclusive of testing surface- and bulk-softness of sheets could respond the perceptions of users; The reciprocal matrix using paired comparison method to the n (n - 1) / 2 power was found to be capable of compensating for the biases of testers and samples and generating a more-reasonable and -robust ranking, also hand-felt softness of the samples could be effectively quantified. Three phases of trial were conducted. The first phase, seven merchandized facial tissue were collected from marketplaces for testing, three major results were achieved: 1) Hand-felt softness inference method established by reciprocal matrix was approved to meet paper manufacturing theories by two panels, experts and laymen. 2) The correlation between two quantified rules established separately by experts and laymen was R2 = 0.985. 3) The proposed pre-test training materials and method were found to be able to enhance the perceptive abilities of the examiners. The second phase, four grades facial products were collected from production lines for testing, four major results were achieved: 1) Triangle test and reciprocal matrix approach could be used together to identify and train examiners. 2) Sheet tensile strength including MDT, CDT and TTS were major contribution to hand-felt softness approved by designed samples. 3) Consistency analysing and adjusting could eliminate ranking contradiction. 4) Adjusting trends and hand-felt samples could be effectively reviewed together by examiners. The third phase, three trail sections were conducted based on the findings from previous two phases, two results were found: 1) In order to calibrate the data of hand-felt softness from two or more panels, more than three standard samples ( n ≥ 4) were required for an optimal liner regression. 2) Trend chart of inconsistence adjustment approaches could be employed to realize the robustness of hand-felt ranking, which was better than using marginal-weight calculated in sensitive analysis. To conclude, the reciprocal matrix and inference method using in this study to quantify hand-felt softness of facial tissue is consistent with paper manufacturing theories approved mathematically that also contribute a procedure from data collection to process feedback, three major processes have been established : 1) Sample preparation and hand-felt softness testing process. 2) Quantification and inference process for hand-felt softness. 3) Linear ability analysis and application process. These three processes can be employed in the quality managements, product developments and marketing surveys.