Community Research and Development Information Service - CORDIS


veggiEAT Report Summary

Project ID: 612326
Funded under: FP7-PEOPLE
Country: United Kingdom

Periodic Report Summary 1 - VEGGIEAT (To critically evaluate vegetable acceptability through individual and environmental characteristics across the lifespan in institutional food service)

Vegetables are one of the most difficult categories of food to introduce into a diet especially within a foodservice operation. The product attributes, the individual characteristics of the consumer and the eating environment all play a key role in food-related decisions. The aim of VeggiEAT is to develop an EU platform for predictive modeling of processed vegetable intake that takes into account individual characteristics (acceptability, intake level, age groups) as well as environmental cues (choice architecture and institutional setting). This aim will be achieved through the development of consumer-oriented products (sensory evaluations); the development of recipes for use by institutional food providers (restaurants, canteens, etc.); and the benchmarking of choice architecture facilitating the consumption of vegetables.

The objective of WP2 was to evaluate the sensory characteristics of the vegetables that would influence their choice by different age groups. The specific objectives were from a methodological perspective: To optimize the free sorting task in regards to the kind of products (i.e. vegetables) and to the kind of consumers (i.e. age / nationality) and from a knowledge perspective: To better understand consumer perception of the product sensory variations according to their characteristics (i.e. age, gender, nationality) leading to operational output: Key information/ recommendations for product range rationalisation and design. Furthermore this WP addressed the following Industry challenges: Sensory product characterisation; Perception of sensory variation and acceptability according to consumer characteristics; Optimisation of sensory and consumer tests according to subject characteristics and suitable for the specific vegetable products (from a technical perspective).
Research components of WP2 are: Task 2.1 (product characterisation): Consists of the sensory characterisation of vegetables by means of descriptive methods. Task 2.2 (consumer test design): Optimisation of several parameters of the free sorting task method. Task 2.3 (consumer tests): Evaluation of sensory variation discrimination according to consumer background (age, nationality) applying a free sorting test and collecting liking and questionnaire responses. Methodologies and approaches in planning and conducting research activities in relation to relative aims are introduced for each task.
Canned peas and sweet corn were selected among the large diversity of food in the vegetable category considering their large availability on the market, their different history and use modalities in European culinary tradition. Ten peas (codes: A,B,D,E,F,J,L,O,P,Q) and eight sweet corn (codes: H,R,S,T,U,V,W,Z) samples were selected in order to cover as much as possible the diversity available for each vegetable.
Subjects were recruited in school and elderly care institutions and/or leisure facilities in Copenhagen (Denmark - DK), Lille (France - FR), Florence (Italy, IT) and Bournemouth (United Kingdom, UK). Ethical approval was sought and granted through standard university procedures in all countries. In total 497 adolescents (mean age 14) and 498 elderly (mean age 68.75) were recruited from the four countries.
Product characterization
The sensory profile of tested samples was obtained by means of Descriptive Analysis (DA). Two panels of 12 subjects participated in the analysis of peas and corn, respectively. Twenty-six and twenty attributes describing appearance, aroma (odour by nose), flavour and mouthfeel characteristics were defined for peas and for corn, respectively. For each product, samples were evaluated in triplicate. Intensities were rated on a 9-point category scale.
Products were also instrumentally characterized.

Consumer test: Experimental procedure
The experimental procedure consisted of three steps: 1. Liking task, 2. Collection of Questionnaire data; 3. Sorting task.
- Liking task: Participants were provided with individual trays with coded pea or sweet corn samples. Subjects were asked to look at the appearance, smell and taste a tea-spoon of each sample, then they were asked to rate their liking on a 9-point category scale.
- Questionnaire: After completing the liking task, subjects filled in a questionnaire consisting of 3 sections: 1. socio-demographic background (age, gender and education); 2. stated liking for a list of eleven vegetables widespread in Europe on a 9 point category scale (1:dislike extremely; 9:extremely like) ; 3. familiarity with the same vegetable list on a 5 points category scale.
- Sorting task: In the last part of the session subjects were provided with a new tray with 11 or 9 three-digit coded pea or sweet corn samples. Subjects were asked to observe, smell and taste samples and then to group those according to their similarities, the number of groups formed should be no less than 2. Subjects were asked to take note on individual ballots of their own criteria used to group samples.

Results and conclusions
The activities related to Task 2.1 in WP2 provided a detailed sensory description of canned peas and sweet corn samples commonly available in the market. Main sensory differences among samples were identified for both products. This information is essential for achieving two aims: exploring sensory characteristics driving elderly and adolescents liking across Europe and studying the relationship between sensory and instrumental data to improve the quality control of these products. In the present study DA provided: 1) a validated sensory profile of each sample, 2) the relative importance of appearance, flavour and texture attributes in discriminating products by means of perceptual maps. The study of the relationship between sensory properties and instrumental measurements was then possible. The projection of Firmness and NMR data onto the obtained sensory spaces resulted in a good evidences of the potential use of these measurements to predict relevant sensory differences among samples.
Results from activities related to Task 2.2 showed that the minimum number of consumers required for free sorting studies seems higher than that recommended in previous works in which product configurations were considered stable when working with more than 25-30 consumers. For both adolescent and elderly a minimum number of 50 subjects is fair when working with familiar canned vegetables such as peas. A larger panel size (70 or more) is required when working with less familiar products such as sweet corn. However our results showed that product knowledge and in particular one of its components (respondent familiarity with the product under investigation) is a factor that should always be considered to define the panel size to perform a sorting task.
Results from the consumer study (Task 2.3) showed that both elderly and adolescents are able to sort vegetable samples in relation to sensory properties that are relevant for their hedonic judgment about the product. High correlation values were found in comparing sorting configurations from each country and each age group with the perceptual maps from descriptive analysis for both peas and sweet corn. Sample grouping was consistent across countries with minor differences that seem to be related to the degree of familiarity of the product in a country. When the sorting task is conducted with familiar products (like peas) differences among countries and age groups tend to be minimal. Both elderly and adolescents showed no difficulties in eliciting terms (sensory and hedonic) that describe the characteristics of the groups they formed in the sorting task. This means that this approach is an effective method to explore vegetable perception in both age groups and obtain information about sensory and hedonic dimensions driving product discrimination. When applied in cross country and across age studies, the free sorting task overcomes limitations of other approaches (e.g. rating method and questionnaires) in which results might be strongly affected by cultural differences in the expression of results (e.g. differences in the use of rating scale across countries and ages).
Lists of terms of perceived properties of peas and sweet corn samples were obtained for the two age groups from all countries. This output is relevant when the interest is focused consumer language in order to better understand sensory barriers to increase vegetable consumption. Relative differences were found in the number and nature of terms used to describe sample groups formed during the sorting task across countries and ages. Appearance seems to be less relevant for elderly than for adolescents in discriminating samples. Older respondents tend to focus their attention more on texture and hedonic terms. The juxtaposition sweet vs bitter; richness in flavour vs lack of taste, always associated with hedonic terms drove product discrimination independently from countries and age groups. The study of the correlation between the occurrences of consumer terms and intensity data from descriptive analysis allowed to “translate” consumer language in sensory characteristics. For instance the term “bad taste” was found to be associated with more technical sensory attributes like “acrid” or “metallic”. Similarly the generic negative hedonic expression “bad texture” was found to be associated with the sensory attribute “softness” in sweet corn samples and “hardness” in pea samples. This information is of great importance to set up proper quality control in food companies.
Task 2.3 also explored and compared the actual liking of adolescents and elderly across the four European countries. Results confirm the effect of familiarity on stated and actual liking for vegetables. The more familiar respondents are with a specific food, the more they will like and prefer it. In the present study the more familiar the respondents were with a vegetable the higher were the differences in liking among the presented samples. For instance, French and Italian adolescents were more familiar with and expressed a higher stated liking for peas than for sweet corn. An opposite trend was observed for Danish and British adolescents. As a consequence British and Danish teens scored their liking for sweet corn samples significantly higher than for pea samples.
The analysis of individual differences in liking allowed us to understand the role of flavour and texture in canned peas and sweet corn acceptance from actual tasting experimental sets. The within-product approach used in this study highlighted that, independently from familiarity and stated liking, main drivers of actual liking and disliking are the same across countries and ages. Sweetness, in opposition to bitterness and sourness, confirmed to drive actual liking for vegetables. The influence of saltiness on liking was positive for peas but negative for sweet corn. Similarly softness was positively related to liking for peas and negatively for sweet corn. Richness in flavour and in colour was strongly correlated to liking for both peas and sweet corn. This information should be taken into account by food producers and the catering sector when promoting the consumption of peas and sweet corn among adolescents in Europe. In relation to VeggiEAT research project the results of WP2 feeds WP3 where recipe development is complete and evaluation is underway led by the Institute Paul Bocuse Research Centre.

Please see attachments for project logo and photographs illustrating and promoting the work of the project. –


Paul Lynch, (Senior Research Support Officer)
Tel.: +44 1202 968265
Fax: +44 1202 965237


Life Sciences
Record Number: 189034 / Last updated on: 2016-09-19