The Research Of Assessment Technique On The Sensory Panel And Panelist And System Development

Sensory evaluation is a scientific method,which is used to evoke, measure, analysis and interpret products through sensory characteristics perceived by sight, smell, touch, taste and hearing and. Currently, sensory evaluation has been widely used in various industries, such as product development, quality control, product optimization, product classification and other aspects. As the instrument of sensory evaluation, panels and panelists are the key to obtain reliable sensory date, and assessment technique on the panels and panelists is an important way to manger this instrument.According to the present research, consistency, repeatability and reproducibility is not only a key index to assess the performance of panel and panelist, but also the major factory to establish a rational, comprehensive assessing system on the discrimination, ranking, quantitative descriptive analysis capability in this paper.In this paper, the major results and conclusions are summarized as follows:(1) For the discrimination capability, the assessment method is based on binomial data. Combining binomial distribution statistics with Guessing Model, we can calculate their PC, PD, PD’RSD, etc. to represent panel and its members’ consistency, repeatability, so that evaluate their each discrimination performance.(2) For the ranking ability assessment, the assessing system is built on rank data. The consistency of panelist can be evaluated by Spearman correlation test and Kendall test examines the repeatability of theirs. An assessment system for panelist is built to check their ability’s correctness and stability comprehensively. While both Page test and LSD are used to achieve panel’s result. The panel is treated as a special panelist and then monitored. The ranking ability system has better reasonable, wider applicability, which improves the existing studies considered only a single indicator.(3) For the quantitative descriptive analysis capability, assessing system is built based on quantitative data. By the intermediate variable of Analysis of Variance, such as the arithmetic mean, mean square error, with standard deviation, etc., also F-test, it helps us to built panelist’s quantitative descriptive analysis capability based on one-factor analysis of variance while panel’s based on two-factor analysis of variance. We can examine consistency, repeatability, reproducibility and monitor their each quantitative descriptive analysis performance.(4) After the establishment of a complete evaluation system, C# programming language is used to develop software by Microsoft Visual Studio tools, advancing with the current information era.In addition, this paper takes a ranking experiment as an example. We organize a sour ranking experiment then applying the assessment system and evaluation software practically. The data by the experiment also is analyzed to verify the applicability of the assessment system, the consistency between the assessment system and the evaluation software, the correctness of the evaluation software. We let twelve panelists join the ranking experiment, using Spearman test and Kendall test to analyze their consistency and repeatability. The analyzed result shows the same conclusion comparing with their experiment result. Then, we combine our assessment system with clustering analysis to classify 12 panelists’ performance, whose sour ranking ability is classified into 3 categories. This result is completely consistent with the scatter diagram for sour ranking ability. Above all that reflects the better applicability and scientific of our research.