Relating consumer and trained panels’ discriminative sensitivities using vanilla flavored ice cream as a medium

This research investigated the possibility of uncovering a relationship between the sensitivities of trained and consumer panels in three experiments. In Experiment I, five studies were conducted using vanilla flavored ice cream. In each study, two ice cream samples differing in formulation and/or their manufacturing process were used. They were compared by both panels using same–different tests with sureness judgments (degree of difference methodology). Using the appropriate probabilistic Thurstonian model, d′ values, a measure of the underlying sensory difference perceived between the products, were calculated and the underlying relationship between the two panels’ sensitivities uncovered. An additional study was then conducted (Experiment II). A new pair of ice creams differing in fat content was first evaluated by the trained panel. Based on the estimated d′ value (trained panel’s measured d′ = 2.69), the corresponding consumer d′ value was predicted using the relationship uncovered in Part I (consumer panel’s predicted d′ = 1.54). The same pair of ice creams was then evaluated by the consumer panel. The measured d′ value for the consumers was almost identical to that predicted by the uncovered relationship (consumer panel’s measured d′ = 1.56). Along with the discrimination component of these studies, paired preference tests were conducted (Experiment III) in order to study a link between perceived difference and expressed preference. The results give an indication of when a perceived difference might start translating into a change in acceptability of the original product. These results indicate the potential of such an approach to predict consumers’ perceptions from an in-house semi-trained or trained panel, providing a useful predictive tool and a means of reducing repetitive and costly consumer testing.     

Sequential use of the triangle, 2-AC, 2-AFC,and same-different methods applied to a cost-reduction effort: consumer learning acquired throughout testing and influence on preference judgements

Taste discrimination and preference of loyal, heavy users of a popular fruit beverage were measured to support a cost reduction effort. The research objective was to determine the highest reduction level permissible without effecting a noticeable perceptual difference among heavy users of the beverage. Judgements of an internal employee panel were considered for predictions of loyal, heavy-user consumer response. Heavy users were defined as individuals who consumed the beverage at least once a week or more. Internal panel 2-AFC and consumer triangle, 2-AC, 2-AFC, and same-different discrimination results collected using the sip methodology were compared using d′ — a measure of the degree of sensory difference to: (1) support the cost reduction effort; (2) study the relationship between scale methods; and (3) examine and establish the relationship between internal and external consumer sip test results. Results of these combined studies showed that while the ingredient reduction was perceived as different even at the lowest levels of reduction and delivered negative implications at 100% reduction, a mid-reduction level was successful in satisfying consumer preference. In light of these findings, the action-decision criteria for moving forward with the cost-reduction reformulation was modified from ‘no perceivable difference’ to ‘parity or superior preference’. Results of these studies led to a successful cost-reduction of the beverage.     

Corroborating the 2-AFC and 2-AC Thurstonian models using both a model system and sparkling water

Two experiments were conducted to investigate the suitability of the Thurstonian models for the 2-AFC (2-alternative forced choice) and 2-AC protocols (2-AFC test with a “no difference” option). The Thurstonian approach predicts that if the same subjects and products are used, the degree of difference measured between the products (d^′) should be the same. Experiment I tested this prediction using solutions with different salt concentration. The average d^′ values for the 2-AFC and 2-AC were found to be 1.66 and 1.54, respectively, and were not significantly different (t-test, p=0.46). This indicated the suitability of the models for a taste stimulus. Experiment II used a trigeminal stimulus: sparkling water with different CO₂ levels as a food system. Three levels of carbonation were used and all three possible pairs of samples were compared. The d^′ values measured for the 2-AFC and 2-AC for each pair of products were: pair 1––1.7 and 1.7; pair 2––0.8 and 0.7; pair 3––1.2 and 1.0. For each pair of products, no significant difference was observed between the d^′ of each protocol (t-tests, p=0.65, 0.20, 0.30, respectively), giving further support to the Thurstonian models of these paradigms and indicating their robustness.     

Observed discriminability is more variable than predicted by signal detection theory

Food scientists and technologists are interested in how sensitive judges and consumers are to changes in product formulations. Numerous approaches to measuring sensitivity have developed at a rapid rate in the last 25 years, however, the evaluation and assessment of sensory tasks is still ongoing. The current study compared the efficiency of four difference tests (A-Not A, A-Not AR, 2-AFC, 2-AFCR) across three concentrations (difficulty levels) of orange essence. To determine which of the four tasks had the lowest variance in d′, the decision strategy used by the judges first had to be ascertained. Using the rating-scale approach of signal detection theory, four estimates of sensitivity, d′, were obtained for each judge (n = 10) at each of the three difficulty levels, for all four tasks. Results indicate that the majority of judges adopted the β-strategy for the A-Not A and A-Not AR tasks, and the β/τ-strategy for the 2-AFC and 2-AFCR tasks. The A-Not A and A-Not AR tasks produced greater estimates of d′ than the other two tasks, and the A-Not A task exhibited a greater variance in d′ than the other three tasks.     

How does it make you feel? A new approach to measuring emotions in food product experience

The objective of this study was to define an approach to describe the emotional profile specific to a product category able to solve some limitations of the current approaches. A multistep approach was applied to measure emotions related to consumer liking for a specific food product category, chocolate and hazelnut spreads, chosen as the first case of application. Three interrelated experiments were conducted: (1) selection of products that spanned the most relevant sensory variation within the considered product category; (2) development of a product specific questionnaire (EmoSemio) based on interviews conducted with a modified version of the Repertory Grid Method and analysed with a semiotic approach; and (3) collection of liking and emotional consumer responses with EmoSemio and with EsSense Profile™ questionnaires. Both questionnaires used to measure emotional responses produced information that is not captured by only measuring acceptability. However results from ANOVA model applied on emotion ratings highlighted important differences between the two questionnaires. The product specific questionnaire was found to discriminate across the products more effectively, with a higher percentage of discriminating emotions and a higher number of sample groups discriminated by each emotions (LSD_99% post hoc test). Different factors contributed to these results: (a) the product-specific and language/culture-specific nature of the questionnaire; (b) a different and clearer way to express emotions in EmoSemio: not using single adjectives but full sentences helps to reduce ambiguity; and (c) a reduced length – 23 instead of 39 items. For these reasons, EmoSemio approach seems to be appropriate when the emotional profile of a specific product category is of interest, allowing a fine-grained analysis with relatively modest costs as to the benefits (25 interviews). Further studies are needed to experiment EmoSemio on other product categories, testing its reliability and suitability with different food and also non-food products.     

Delta-tilde interpretation of standard linear mixed model results

We utilize the close link between Cohen’s d, the effect size in an ANOVA framework, and the Thurstonian (Signal detection) d-prime to suggest better visualizations and interpretations of standard sensory and consumer data mixed model ANOVA results. The basic and straightforward idea is to interpret effects relative to the residual error and to choose the proper effect size measure. For multi-attribute bar plots of F-statistics this amounts, in balanced settings, to a simple transformation of the bar heights to get them transformed into depicting what can be seen as approximately the average pairwise d-primes between products. For extensions of such multi-attribute bar plots into more complex models, similar transformations are suggested and become more important as the transformation depends on the number of observations within factor levels, and hence makes bar heights better comparable for factors with differences in number of levels. For mixed models, where in general the relevant error terms for the fixed effects are not the pure residual error, it is suggested to base the d-prime-like interpretation on the residual error. The methods are illustrated on a multifactorial sensory profile data set and compared to actual d-prime calculations based on Thurstonian regression modeling through the ordinal package. For more challenging cases we offer a generic “plug-in” implementation of a version of the method as part of the R-package SensMixed. We discuss and clarify the bias mechanisms inherently challenging effect size measure estimates in ANOVA settings.     

Modelling of rheological,microbiological and acidification properties of a fermented milk product containing a probiotic strain of Lactobacillus paracasei

The simultaneous effect of fermentation temperature (FT, 36.7–43.4°C), milk total solid level (TS, 11.3–14.7%, w/v) and total inoculum concentration (TI, 2.16–3.84 v/v) on the acidification process and the rheological properties of fermented milk products with Lactobacillus paracasei ssp paracasei B117, Lactobacillus delbrueckii ssp bulgaricus Y 6.15 and Streptococcus thermophilus Y 4.10 was explored by means of response surface methodology. Maximum storage modulus (G′_max), minimum loss tangent (tan δ_min), rate of gelation (I_E) and onset of gelation were the rheological parameters studied. Maximum acidification rate (V_m), time at which maximum acidification rate was observed (T_m), and time to reach the end of fermentation (T_e) characterized the kinetics of acidification, whereas the increase in the number of the three bacteria at the end of fermentation was chosen as the microbiological parameter of the system. The growth/survival of microorganisms and the organic acid profile during cold storage as well as the overall product acceptability by a consumer panel were also assessed. TS strongly affected G′_max and tan δ_min; high TS resulted in large increase in G′_max and decrease in tan δ_min. Increasing fermentation temperature gave a decrease in the onset of gelation, V_m, T_m and T_e, and an increase in the gelation rate (I_E). Under conditions of relatively low FT (37–40°C), high TS (about 14%) and high TI (3–4%), relatively high gelation and low acidification rates were observed, fermentation took a longer time to finish, but the formed gels were firmer, showing higher G′_max and lower tan δ_min values. Low FT (36–38°C) enabled higher increase in the number of L. paracasei B 117. The probiotic strain showed good compatibility with the S. thermophilus Y 4.10 and L. bulgaricus Y 6.15, and satisfactory levels of all bacteria were found during fermentation and storage at 4°C for 21 days. No major differences in lactic and uric acid contents were seen between the control (probiotic strain-free product) and the probiotic fermented milk, whereas the latter contained slightly higher amounts of citric, pyruvic and orotic acids. Moreover, the probiotic fermented milk was graded by the consumer panel with a similar acceptability score as the control product.     

The effect of consumption context on consumer hedonics,emotional response and beer choice

The context of a consumer test affects participant response. Data collected in a sensory laboratory is likely to have little predictive value of consumer experience in real-life situations. This study determined the effects of context on consumer response to two commercial beers. Regular beer consumers (n = 100) rated liking and emotional response using ten beer-specific emotion categories for two beers (Lager and Ale) under three different conditions: (1) a sensory testing facility (Lab), (2) a natural consumption environment (Bar) and (3) using an evoked context (Evoked). Their choice of product to take home was also recorded. Overall results showed significant product differentiation for liking (F (99, 2, 1) = 8.46, p = 0.004) and product choice (Q (1, N = 100) = 4.85, p = 0.028) in the Bar but not in the Lab or Evoked context. Emotional variables highlighted significant product differentiation (p < 0.05) but more so in the Bar than in the Lab or Evoked context. However, clustering participants on liking revealed three distinct clusters differing in sensitivity to context. Two clusters showed opposing but consistent preference for one of the two products regardless of context. The third cluster was more influenced by context, showing a more discriminating response in the Bar. These findings showed that consumers differ in their degree of context-sensitivity and the extent to which evoking a context gives similar results to a real environment. They also highlighted the importance of segmentation and confirmed the added insights gained by measuring emotional response compared to liking.     

Effects of an evoked refreshing consumption context on hedonic responses to apple juice measured using best–worst scaling and the 9-pt hedonic category scale

Awareness of the need to consider a product’s consumption context when measuring consumer hedonic response of a product is increasing among consumer sensory researchers. This study investigated the effects of evoking a consumption context using a written scenario on hedonic response measured using best–worst scaling and the 9-pt hedonic category scale. Hedonic responses for four apple juices with relatively large sensory differences were compared when measured in the evoked context ‘when having something refreshing to drink’ using best–worst hedonic scaling (n = 65) and the 9-pt hedonic scale (n = 48). Best–worst scaling discriminated between the four apple juices when a refreshing context was evoked (p < 0.01), while the juices were equally liked using the 9-point scale (p = 0.41) when the same context was evoked. Consumers perceived best–worst scaling to be more difficult than the 9-pt scale, however there was no difference between the two methods for consumers perceived accuracy of their liking information. The present study highlights that the effect of an evoked context on hedonic response may not be universal for hedonic methods. Further research is needed to understand the effect of evoking context on the liking of products, and to determine whether this measure reflects product liking in an actual consumption context.