Comparison of methods to monitor liking of food during consumption (original) (raw)
Related papers
Food Research International, 2017
The idea of having untrained consumers performing Temporal Dominance of Sensations (TDS) and dynamic liking in the same session was recently introduced (Thomas, van der Stelt, Prokop, Lawlor, & Schlich, 2016). In the present study, a variation of the data acquisition protocol was done, aiming to record TDS and liking simultaneously on the same screen in a single session during multiple product intakes. This method, called Simultaneous Temporal Drivers of Liking (S-TDL), was used to describe samples of Gouda cheese in an international experiment. To test this idea, consumers from six European countries (n = 667) assessed 4 Gouda cheeses with different ages and fat contents during one sensory evaluation session. Ten sensory attributes and a 9-point hedonic scale were presented simultaneously on the computer screen. While performing TDS, consumers could reassess their liking score as often as they wanted. This new type of sensory data was coded by individual average liking scores while a given attribute was perceived as dominant (Liking While Dominant; LWD). Although significant differences in preference were observed among countries, there were global preferences for a longer dominance of melting, fatty and tender textures. The cheese flavour attribute was the best positive TDL, whereas bitter was a strong negative TDL. A cluster analysis of the 667 consumers identified three significant liking clusters, each with different most and least preferred samples. For the TDL computation by cluster, significant specific TDL were observed. These results showed the importance of overall liking segmentation before TDL analysis to determine which attributes should have a longer dominance duration in order to please specific consumer targets.
2014
Generally liking is measured overall but is likely to vary over the food intake, like sensory perception. Replacing the attributes in Temporal Dominance of Sensations (TDS) by the categories of a usual ordinal liking scale makes it possible to monitor liking changes while eating a product (Sudre et al., 2012). This methodology allows for a better understanding of the influence of temporal dominance of sensations on liking and liking evolution over the time of product intake. Thus, it is possible to associate hedonic temporal data and descriptive temporal data (TDS profiles), which would identify drivers of liking, that is attributes which, when cited as dominant, would lead to a decrease or an increase of liking. Sixty-eight French consumers participated in four lab sessions. During the 1st session, the subjects had to taste six flavored fresh cheeses and to indicate their overall liking for each product. Temporal liking of the same products was evaluated throughout the tasting in the 2nd session. The 3rd session was a measure of TDS for the same products and only dominance was recorded (no intensity recorded). Finally, in the 4th session, the measure of liking of the same products was recorded 1 min after the beginning of tasting of each sample. Measures of liking showed that temporal liking is more discriminative than classic liking, and that classic liking scores may have been given before swallowing. However, waiting for 1 min before asking a liking score decreases discrimination of the products. Furthermore, combining temporal liking and TDS data obtained from the same consumers allowed to suggest the attributes and the times to which the perception of those attributes is responsible for the liking or disliking of the product. This is a new concept of ''positive or negative Temporal Drivers of Liking'' (TDL). The paper proposes to compute individual average liking scores while a given attribute was perceived as dominant (LWD) and to average those LWD values over the subject for extracting the TDLs of a specific product or of a group of products.
The contribution of sensory liking to overall liking: An analysis of six food categories
Food Quality and Preference, 1995
Individuals differ in the degree to which they attend to different sensory inputs of food when they make their judgments of overall liking. Foods are complex, so it is quite difficult to isolate the effect of any one sensory input without confounding with other sensory inputs. However, analytically one can begin to estimate the importance of a sensory input by the linear slope M, relating overall liking to sensory attribute liking [Overall Liking = M(Attribute Liking) + B]. On an individual-by-individual basis one can compute these slopes for appearance, taste/flavor, and texture, respectively, as well as the normalized slopes (defined as the attribute liking slope divided by the sum of the three attribute liking slopes). The approach is applied to six food product categories: bologna, hot dog, carbonated fruit beverage, blueberry pie filling, peanut butter, salad dressing. The results show substantial inter-individual variability in the way overall liking covaries with sensory attribute liking. On an aggregate basis for foods the rank of importance for attribute liking is taste/flavor, texture, and appearance, respectively. However, on an individual basis there are a variety of different patterns.
Dynamic aspects of liking: post-prandial persistence of sensory specific satiety
Food Quality and Preference, 2005
The time dependence of the liking of foodstuffs was investigated in a study with 25 subjects, consisting of three parts: (1) on day one, a sensory specific satiety study, which was extended until 125 min after consumption; (2) on days two to seven, a repeated in home taste and evaluation study; (3) on day eight, a second sensory specific satiety study as in 1.
PLoS ONE, 2014
Consumers show high interindividual variability in food liking during repeated exposure. To investigate consumer liking during repeated exposure, data is often interpreted on a product level by averaging results over all consumers. However, a single product may elicit inconsistent behaviors in consumers; averaging will mix and hide possible subgroups of consumer behaviors, leading to a misinterpretation of the results. To deal with the variability in consumer liking, we propose to use clustering on data from consumer-product combinations to investigate the nature of the behavioral differences within the complete dataset. The resulting behavioral clusters can then be used to describe product acceptance. To test this approach we used two independent data sets in which young adults were repeatedly exposed to drinks and snacks, respectively. We found that five typical consumer behaviors existed in both datasets. These behaviors differed both in the average level of liking as well as its temporal dynamics. By investigating the distribution of a single product across typical consumer behaviors, we provide more precise insight in how consumers divide in subgroups based on their product liking (i.e. product modality). This work shows that taking into account and using interindividual differences can unveil information about product acceptance that would otherwise be ignored.
Foods, 2018
Conventional drivers of liking analysis was extended with a time dimension into temporal drivers of liking (TDOL) based on functional data analysis methodology and non-additive models for multiple-attribute time-intensity (MATI) data. The non-additive models, which consider both direct effects and interaction effects of attributes to consumer overall liking, include Choquet integral and fuzzy measure in the multi-criteria decision-making, and linear regression based on variance decomposition. Dynamics of TDOL, i.e., the derivatives of the relative importance functional curves were also explored. Well-established R packages 'fda', 'kappalab' and 'relaimpo' were used in the paper for developing TDOL. Applied use of these methods shows that the relative importance of MATI curves offers insights for understanding the temporal aspects of consumer liking for fruit chews.
Dynamics of food preferences: a case study with chewing gums
Food Quality and Preference, 2004
The dynamics of food perception is critical in sensory science. However, although much research deals with time-intensity measurement as a tool for analytical sensory evaluation, it is striking that almost no published research investigates the dynamics of hedonic responses. Here we studied how consumers scored mint flavored sugar-free gums when the test duration varied. Six coated drag ee-like gums from the French market were tested by each consumer according to three test conditions: 1 min of chewing, 5 min and half-an-hour. One and five minute tests took place in the sensory lab evaluation booths whereas consumers were free to move about for the 30 min modality. All consumers thus participated to a total of 18 sessions. Our main finding is that the average liking varies with the test duration but that this variation differs for the six products in a way that one of the initially least liked products becomes the most liked when it is chewed during 30 min. Also, the individual liking patterns and the resulting consumer segmentation varied widely across the three test conditions. Additionally, a Flash profile was performed with experienced subjects in an attempt to relate the observed differences in liking to the gum sensory characteristics. Trigeminal-related sensations seem to play an important role in these changes. Clearly such aspects are to be taken into account for new product development and market analysis. This kind of concerns may be even more challenging for hedonic evaluation of many non-food products.
Journal of Sensory Studies, 2008
An alternative rating method, known as positional relative rating (PRR), or rank rating, was compared with the traditional 9-point hedonic scale. Initially, category scale and line scale versions of PRR were investigated to see which of either was better at differentiating samples. Four lemonades were assessed in duplicate on both scales (category and line). Overall quality, sweetness intensity and sourness intensity were assessed in separate sessions by groups of~50 consumers, depending on session. The category scale better differentiated samples than did the line scale for two of three attributes assessed, with differentiation being the same in the third. Subsequently, the overall liking of these same lemonades was assessed by 100 consumers on both the PRR category scale and the traditional 9-point hedonic scale. The differentiation of samples for both scales was equivalent, although mean values were somewhat lower with PRR.
Food Quality and Preference, 2004
Preference for 7 orange juices was recorded monadically. Three of the samples were replicates of the same juice (RJ). The 4 other samples were RJ slightly spiked with either sucrose, or citric acid, or quinine or an orange flavor. Each subject then performed a paired preference test composed of RJ and the modified juice whose score was the furthest from the RJ score. Finally, subjects were asked to choose one among 8 reasons why they preferred that sample. A number of individual statistics were computed in order to compare the consistency of liking scores within session, across ages and between genders. Women were slightly more consistent than men. However, no significant effect of age nor gender was found in the coherence between the first hedonic test and the paired preference test. Moreover, children and seniors found it more difficult than other consumers to identify the reason for their preference.
Temporal dominance of emotions: Measuring dynamics of food-related emotions during consumption
Food Quality and Preference, 2014
Mapping food-evoked emotions in addition to sensory profiling is topical. In sensory profiling, the Temporal Dominance of Sensation (TDS) method focuses on the assessment of the temporal evolution of dominant sensory attributes over time. We hypothesize that food-evoked emotions also show temporal dynamics that can be related to dynamic sensory perception. This study assessed temporal dynamics of sensory and emotional attributes during chocolate tasting. We used TDS to determine dynamic sensory properties of dark chocolates providing a list of 10 sensory attributes. Comparably, Temporal Dominance of Emotions (TDE) was assessed by replacing the sensory attributes with 10 emotional attributes. Sixtytwo participants assessed TDS and TDE of five commercially available dark chocolates (plain and flavoured). Multivariate comparisons (Hotelling test) showed significant differences between products based on the dominance duration of sensory (p < 0.05) and emotional attributes (p < 0.05). TDS difference curves revealed products to differ based on their dominant sensory attributes, with different attributes peaking at different time moments. TDE difference curves showed that products also differed in the temporal distribution of dominant emotional attributes. Comparing the average dominance rates between plain dark and flavoured dark chocolates revealed that for flavoured dark chocolates mainly flavour attributes and positive/active emotions were perceived as salient whereas for plain dark chocolates textural as well as taste attributes were dominant accompanied by more negative/non-energetic emotions. A joint CVA plot on the duration of dominance for sensory and emotional attributes per product revealed that temporal evolution of sensory -and emotional attributes was related. This suggests a mutual reciprocity between those two entities (sensory and emotional attributes) resulting in more complex, richer product characterization. In conclusion, these findings show TDE to be a promising new venue in characterising food-evoked emotions in relation to sensory profiling.