Special designs of popularity tests, consumer preferences of coffee beverages

Friday, August 26, 2016

Author: Ildikó Bálint

1. Summary

In my work, the balanced incomplete block (BIB) design, common in sensory analysis, but rarely used in product development or the design of experiments, is presented, with the examples of coffee beverages made of different coffee blends. It is a common feature of product development that a large number of consumer assessments (more than sixty assessments per target group, per cell) have to be performed simultaneously regarding many products (more than 6), however the limitations of this are discussed extensively in the literature. So, the reliability of the tests is significantly distorted by sensory fatigue, excessive mental strain and loss of motivation of the assessor, however, these factors can be avoided by developing the appropriate experimental design. One of the major advantages of the BIB design is that only a small portion of the samples (no more than six) are evaluated by the assessors, however, the nature of the result obtained after summarizing and analyzing the assessments is completely in line with the value which would have been obtained if every sample had been assessed by each assessor. Because of the complex nature of the coffee beverages and their fast-changing sensory characteristics, only a total of 4 samples of the 10 different coffee beverages involved in the test were evaluated by each assessor.

2. Introduction

2.1. Consumer sensory testing
According to their qualifications, sensory assessors can be divided into three categories: non-professional (naive) consumer assessors, trained assessors, and expert assessors. For different types of tasks, the application of assessors with different qualifications are required [25], [27]. It is characteristic of the attitudes of non-professional consumers that they experience their sensations, and not analyze them, they rely on their own experience during the assessment, and project their own tastes onto the products assessed. Therefore, questions posed to non-professional consumers are aimed at popularity, preference. During studies, that are called consumer tests in general, a query with a large number of people (at least 60 persons per target group, per cell) representing the basic population (in terms of gender, age, address, level of education, net salary, etc.) based on a sampling plan is performed. Assessors typically do not have prior knowledge of the product, only a few products are evaluated by them, with the help of simplified scales and easy-to-understand, short questionnaires. In this case, in reality, the personal, subjective taste is analyzed [23], [38], [13].

 read more ...