Over the last decade an impressive amount of evidence indicates that most human behaviour is largely guided by an ‘adaptive unconscious’ that effortlessly directs our decisions. This implies that we often do not really know why we perform certain behaviour and that, when asked ‘why’ we come up with explanatory constructions that are based on ‘shared causal theories’. In fact, the factor past experience usually explains more of the variance in behaviour than hypothetical constructs such as attitudes and beliefs. This means that memory must play a very important role in actual food choice decisions. This memory is based on non-intentional (also called incidental or unconscious) learning and remains implicit knowledge that expresses itself in the unconscious expectations people have about food.

One of the main functions of food memory is to recall previous experiences associated with food. Accordingly, food memory influences food perception during its consumption through the generation of expectancies and the expression of cognitive associations. In view of this fundamental role, recent studies focused on obtaining a better insight of food memory properties. In this research indications have been obtained that methods based on food memory may be more important for the measurement and prediction food appreciation than the traditional methods based on perception. Memory integrates sensory input, filters out unimportant and retains essential features of the food that are unconsciously perceived but are later missed when not found. This workshop is designed to encourage lateral thinking and information exchange of ideas about food memory properties and the implications of such properties on food and consumer science. The workshop will consist of a mixture of question focused presentations and moderated discussions followed by feedback, final discussions and conclusions.

Keywords: memory, learning, consumer

After a brief introduction and three short presentations (please, find the corresponding abstracts below), workshop participants will be divided into three groups and encouraged to discuss the presentations and to contribute their ideas based on experience. Individuals from member institutes of the European Sensory Network, ESN, will act as group moderators.

1. The role of food memory in everyday life

Jos Mojet and Ep Köster, Centre for Innovative Consumer Studies Wageningen University and Research, The Netherlands

Recent research has used a new paradigm that permits the study of food memory and memory for food related senses without disturbing the incidental learning process that takes place in natural eating situations. It differs from traditional recognition paradigms in a number of respects. People remain unaware of the fact that they learn something, the target foods ar not itemised and presented among other items, but remain embedded in a normal eating situation and the test consists of  recognising the target food among small variations of that same food, instead of among completely different food items. Furthermore, the memory tests are either „absolute“ (Did youhave this during the meal?) or „relative“ (do you like this one more, less or as well as the one you had in the meal?) The results show that people are guided by feelings of „not knowing“ in making their judgments, rather than by theprecise  remembrance of  the previously encountered and incidentally learned food. It is also clear that this form  incidentally learned memory is remarkably resistant to losses with age and that this form of perceptual memory may be different for olfaction and for vision. The experiments have shown a number of instances of memory selectivity and memory distortion that raised questions about the validity of perceptual measurements as a basis for food launching decisions. Consumers make their decisions on the basis of their remembrance of foods and not on the basis of their actual perceptions.

• Central theme for discussion group 1: Would memory be a better measure and predictor of consumer satisfaction and the emotional value of foods than perception? (Moderator: Anne Goldman, ACCE, Mississauga, Ontario, Canada)

2. Food memory research with children

Monica Laureati and Ella Pagliarini, Department of Food Science and Technology and Microbiology, University of Milan, Italy

Recently an implicit paradigm to measure food memory in an ecological way has been proposed and validated. Such a paradigm has been applied to study memory for different sensory modalities and to investigate the effect of aging on food memory performance.

It is largely believed that memory deteriorates with age. There is evidence that the loss of ability to recall previously encountered experiences is particularly evident for explicit memory, whereas implicit memory is relatively unaffected by age. Accordingly, results on food memory performance in consumers of different age indicate that adults and elderly people have comparable memory indices, thus suggesting that food memory is resistant over the lifespan.

Very little is known about food memory in children. The lack of information on children food memory is especially due to the particular attention and the specific methodologies that children need according to their skills and cognitive ability.

It has been suggested that teens older than 15 years of age are generally capable of performing sensory tests like adults. However, with pre-schoolers, early readers and pre-teens alternative methodologies, such as appropriate scales and instructions are recommended.

In food memory research with children, absolute and relative memory are usually assessed using the same approach than for the adults. These studies indicate that children are able to learn incidentally the food they eat and show absolute and relative memory abilities comparable to that of the adults. The outcome has interesting implications in children’s eating and drinking behaviour.

• Central theme for discussion group 2: Is food memory affected by age? Are the methods used to measure memory and hedonic perception applicable with children? (Moderator: Annika Ǻstrom, SIK, Göteborg, Sweden)

3. Memory test versus discrimination test: which is the best to predict consumer feedback?

Léri Morin-Audebrand, University of Applied Sciences Wallis, Rte du Rawyl 64, CH-1950 Sion, Switzerland
Sylvie Issanchou and Claire Sulmont-Rossé, INRA, UMR 1129 FLAVIC, F-21000 Dijon, France

One of the major issues of sensory studies is to understand how consumers make their decisions of food choice and consumption. The classic approach is to consider perceptual measurements as a basis for food launching decisions. For example, in the field of food development, difference tests such as duo-trio, triangle test or A-NonA are usually conducted to check if a change in process or ingredients is perceived or not; a perception of a difference being considered as predictive of a risk of consumers’ rejection. In such difference tests, the ‘old’ version and the new version of the considered food are presented conjointly to panellists to be directly compared. However, in a real life situation, such direct comparison between two products rarely occurs. Instead, the perception of the new product is compared to memorized information from previous experience involving similar foods usually consumed. Consumers make their decisions on the basis of their remembrance of foods and not on the basis of their actual perceptions.

Several questions emerge from this observation: Are difference tests relevant to predict consumers’ reaction, i.e. does a significant discrimination during a difference test necessarily predict that a significant discrimination will occur in a much more natural context of decision about a change in the usual product? Are difference tests and memory tests as much acute for all sensory attributes, i.e. are difference tests in accordance with memory selectivity and memory distortion that are observed in food memory tests?  Thus, this presentation will focus on several studies allowing the comparison between traditional difference tests and memory tests. Distortions between difference and memory tests at individual and group levels will be discussed.

• Central theme for discussion group 3: Should memory tests be preferred to classic difference tests to understand food choice? (Moderator: Sylvie Issanchou, INRA, UMR 1129 FLAVIC, F-21000 Dijon, France).

The ESN sponsored workshop on the "Characteristics of Memory for Foods" elicited much discussion and interest at the Pangborn Symposium in Florence, Italy, September 2009.

The presenters* - experts working in this research area - Ep Koster CICS, Wageningen, Ella Pagliarini, University of Milan and Léri Morin-Audebrand, University of Applied Sciences Wallis, Switzerland introduced the discussion with short presentations that summarised current research knowledge. From the presentations and the following discussion resulted a number of questions and points that present opportunities for ongoing research.

Would memory be a better measure and predictor of consumer satisfaction and the emotional value of foods than our current methods that use analytical perception?

Satisfaction and satiety are independent states that may or may not coincide after consuming a meal or a snack. There is a need to develop a memory based method to measure food satisfaction independently of satiety. Current methods including purchase measures should at least be complemented by memory measurements to try to solve the ongoing problem of new product failures. This will only be demonstrated by correlating memory measures with data from in-market successes. There is evidence that there is enhanced memory for disliked sensory attributes and that learning and memory is not equal for all sensory attributes or product types.

Is food memory affected by age?

Explicit memory shows a loss with age whereas implicit memory is unaffected by age. Results on absolute memory for foods show that adults (18-45 years), elderly people (> 60 years) and children have comparable memory indices. Are the methods used to measure memory and hedonic perception applicable with children? Generally the methods used with adults can be adapted to children’s research, although children are more likely to use extreme ends of the hedonic scales than adults.

Are memory tests better than classic difference tests for understanding food choice?

Difference tests do not replicate real life situations where direct comparisons are rarely made. Instead the perception of a new version of a product is more likely to be compared to memorized information from past experience with the product or with similar products. For the senses of taste, smell and mouthfeel memory is focused on detecting change rather than recognition of the stimuli.  Additionally some people may have good memory but not a good ability to discriminate so that the researcher should consider segmenting individuals based on their respective abilities. Other considerations are the effects of more complex products and the portion or meal size consumed in the test evironment. Any alternative method to difference testing that is based on memory will require followup testing with consumers to understand the predictive ability of the test, including direct observations of what people do.

A final question raised by the discussion:
The paradigm developed to test memory is based on incidental learning (learning without  explicit learning intention), but the memory tests are explicit.

Should we not devise implicit memory measuring tools that bring us even closer to everyday reality ?

* Additional authors:
Claire Sulmont-Rossé, Jos Mojet,  Monica Laureati, Sylvie Issanchou