Context effects in food preference conditioning: cognitive affective and physiological factors.

Lead Research Organisation: University of Manchester
Department Name: Psychological Sciences


This project will examine the affective, cognitive, and physiological mechanisms involved in human food preferences. Food choices are complex behaviours, not yet well understood, and have serious repercussions in terms of public health. We will use the Evaluative Conditioning paradigm (EC) to examine the effect of context on food preference. EC refers to a change in the affective response to a conditioned stimulus (CS) based on its pairing with an unconditioned stimulus (US) (Davey, 1994; De Houwer, Baeyens, & Field, 2005). It is a general mechanism that determines how humans establish likes and dislikes, and may be qualitatively distinct from other forms of (non-affective) learning (Baeyens & De Houwer, 1995). Animal studies show the existence of context effects in food preference (Boakes, Westbrooke, Eliott & Swinbourne, 1997; Hall & Honey, 1989; Maes, Havermans & Vossen, 2000). This may be an important factor in human food choices (e.g., cereals are not eaten for dessert, coffee is not served as a celebratory drink, etc.). Being able to alter these associations may help promote healthier choices (e.g. adopting cereals instead of crisps as a snack). However little empirical research has examined context effects in humans, using either behavioural or physiological studies. For instance, we do not know whether unconditioned rewards (i.e., foods that are naturally liked, such as sugary tastes) are more or less susceptible to the effect of context compared to conditioned rewards (i.e., foods that one learns to like). It is also not known whether associations between food and context can be changed, and how. Answering these questions will improve our understanding of the mechanisms involved in a basic form of affective learning that is likely to be central for food-related behaviour. We will use an evaluative conditioning paradigm where a target food will be positively or neutrally conditioned. We will compare inherently liked and inherently neutral target tastes. As part of the conditioning, we will induce an association between the target food and a context. Context may be defined as a time of day (e.g., breakfast vs. lunch time). This association will be induced either explicitly (through verbal information) or implicitly (through exposure). In a second phase, we will alter the association between the target and the context, through either explicit (information) or implicit (exposure) means. The project will examine the distinction between implicit and explicit processes. An increasing amount of research suggests that implicit affective reactions, evidenced by arousal of the autonomic nervous system (e.g., increased skin conductance, changes in pulse amplitude, etc.) may be better predictors of behaviour than explicit affective evaluations. For instance, cognitive processes may be affected by emotional stimuli that evoke stronger physiological reactions, but not necessarily those stimuli that are explicitly evaluated as being more emotional (Lavda, Blanchette, Richards, & Hamilton, 2006; Blanchette & Leese, 2007; Richards & Blanchette, 2007). In this project, we will measure both participants' explicit evaluations of the conditioned stimuli, as well as measure their autonomic reactions to the presentation of the stimuli. We will compare the relative importance of explicit and implicit responses and examine their relationship to context effects. Thus, the goals of this project are to answer the following questions: 1. Are there contextual effects in evaluative conditioning in humans? (i.e. Can people learn to like a target food in one context and dislike it in another context?). 2. Are explicit or implicit processes more important in establishing associations between food and context? 3. Can the association between a target food and a context be changed, and if so, through what means (explicit or implicit)? 4. Is the strength of the contextual effect different for inherently liked and neutral tastes?


10 25 50