This study proposes the application of ranking to temporal product evaluation to differentiate attribute dominances over time. The ability of the proposed method to discriminate sensory differences was compared to temporal check-all-that-apply (TCATA), a discriminating temporal method for capturing the temporal profile of food and beverage products.
Investigating perception dynamics and uncertainty in temporal sensory data via independent components analysis (ICA)
Wine finish from several Syrah wines are evaluated using temporal check-all-that-apply (TCATA), and results analysed with Independent Components Analysis (ICA) using the Joint Approximate Diagonalization of Eigenmatrices (JADE) algorithm.
Analyzing data using the chance-corrected beta-binomial model: Parameter estimates and their confidence regions
Data arising from replicated sensory discrimination test methods with a correct response are often modelled using the chance-corrected beta-binomial distribution. The model can provide maximum likelihood estimates of the mean proportion of correct responses in the population (and discriminal distances under Thurstonian assumptions) and of the assessor heterogeneity (overdispersion). Both of these parameters are estimated with uncertainty. Previously uncertainty associated with these parameters has been considered only one parameter at a time.
Analysis of sensory check-all-that-apply (CATA) data which includes the evaluation of a single ideal product
When evaluating samples in sensory tests, consumers are sometimes asked not only about real samples but also about imagined ideal products. Check-all-that-apply (CATA) questions are one way to understand consumers’ perceptions of products and their ideal product. We propose the following statistical analyses of consumer CATA data: (i) confidence intervals for head-to-head comparisons based on CATA data; (ii) panel (dis)agreement on whether a sample is characterized in the same way as the ideal product; (iii) contextualizing results via a fragility index; (iv) Monte Carlo tests of independence to determine differences between real and ideal products; (v) the use of mixture of latent trait models with common slope parameters (MCLT) with ideal product data. Continue reading Analysis of sensory check-all-that-apply (CATA) data which includes the evaluation of a single ideal product
A retail simulation study for investigating product choice and choice satisfaction: A case study involving kombucha
Kombucha is a fermented tea that is rapidly growing in popularity in the United States. As part of a larger consumer study conducted in Portland, Oregon, kombucha consumers (n=1303, 68% female, aged 18-86, USA)participated in an online retail simulation test. They evaluated 9 commercial products (bottle images) and made choice selections in a retail simulation.
Does the τ estimate from same-different test data represent a relevant sensory effect size for determining sensory equivalency?
Analysis of data arising from the same-different test method can be submitted to Thurstonian-derived modelling with the goal of estimating the sensory distance between two products (the discriminal distance δ) and the response bias for responding “same” (τ). Previously it has been proposed that it is possible to use τ estimates from same-different test data to represent the consumer-relevant effect size, and thus be used as a criterion for evaluating sensory equivalency, including with other sensory discrimination test methods, such as the tetrad test method. In this paper we explore that proposal.
Segmenting consumers based on food choice motivation and product benefit expectations: A case study involving Kombucha
Kombucha is a fermented beverage that is rapidly growing in popularity in the United States. As part of a larger consumer study conducted in Portland, Oregon, kombucha consumers (n=230) participated in a retail simulation in a central location test. They evaluated 9 commercial products, made choice selections, and were prompted with questions related to food choice motivations, product expectations and choice satisfaction, and other aspects (psychographic, product usage, etc.). We used exploratory multivariate data analyses and cluster analysis methods to explore connections between food choice motivation and product choice, between expectation and satisfaction, and between food choice motivations and product benefit expectations. In this poster we highlight some of the methods used. Results help to better understand consumer expectations and drivers of purchase and repurchase within category-specific consumer segments.
Consumer testing to enable evidence-based business decision-making
By some estimates more than 90% for new food and beverage products fail. In spite of the enormous costs associated with product failures, businesses continue to invest in new product development because the relatively few products that succeed provide business growth and profits. Consumer purchase and repurchase are essential to the success of new products. Consumers are decision makers: by voting with their wallets they determine which products succeed. Consumer input is essential for guiding product development. Consumer feedback can help to identify failure earlier, before more money is invested.
Consumer testing that is grounded in sensory science provides structured data for evidence-based business decision-making. This workshop gives an overview of how a successful consumer testing program can provide critical feedback to the business and to product developers at key stages throughout the product lifecycle. The following topics will be discussed:
- Objectives & fundamentals
- Confirmation of sensory acceptability in scale up
- “Meet-or-beat” testing against benchmarks
- Guidance for product renovations of in-market products
- Data analysis and actionable decision making
Attendees with Internet access (e.g. smartphone) will have an opportunity for completing a hands-on consumer test.
Audience: General
Duration: 2 hours
Continue reading Consumer testing to enable evidence-based business decision-making
Analyzing data from sensory discrimination tests: Parameter estimates and confidence regions
The sensory profile of a food product is an important contributor to its identity, and is linked to brand value. Sensory testing provide an objective way to ensure that the product’s target sensory profile is achieved, e.g., after making production changes. Historical data, including business outcomes, can provide context to assist decision making. Unreplicated sensory discrimination tests without response bias provide simple, actionable outcomes, whereas replicated tests, which are often used to compensate for insufficient resources, allow for estimation of heterogeneity and discriminable sensory distance parameters. Same-different test data can be used to estimate response bias, as well as discriminable sensory distance, and contribute towards an understanding what constitutes a consumer-relevant sensory difference. Usually two-parameter models are interpreted by considering one parameter at a time. This talk shows the potential for jointly investigating multiple parameters via confidence regions. Bivariate analyses can support visualization and communication of results and also provide analytic approaches to enable decision making.
Is less always more?
Here is a question to be answered in mythbusters style: “Do the number of available choices create a choice overload effect in consumers?” As a starting point, consider a well-known study in which consumers receive a coupon in a grocery store after tasting two jams. Consumers who were limited to choosing from only 6 jams followed through with purchase more often than consumers who could choose from 24 jams. On balance, consumers with more options are reported to be less satisfied, less confident, and more regretful of their choice. Results seem aligned with the adage “less is more” and its corollary “more is less.”
Fast forward more than a decade of follow-up studies, some confirming and others disconfirming the choice overload effect. A meta-analysis published in 2013 found no choice overload effect, only high variability in study results. Why is the choice overload effect observed in some studies and not in others? More importantly, when choice overload is realized, what other factors are involved?