| Evaluative conditioning (EC), the pairing of objects (conditioned stimuli, CS) with positively and negatively valenced unconditioned stimuli (US) in order to induce attitude change, has been demonstrated to be a viable method of changing attitudes towards foods and corresponding eating behaviors. Positively conditioning healthy foods and negatively conditioning unhealthy foods should result in healthier food preferences. Of current interest is the extent to which EC can generalize beyond the conditioned foods to entire dimensions underlying food judgment, particularly health and taste. The current research includes three EC experiments using the Video Surveillance paradigm (Jones, Fazio, & Olson, 2009) and three experiments using physical push-pull movements as US. Four healthy CS foods were paired with positive US and four unhealthy CS foods were paired with negative US. Participants then reported eating intentions for a variety of foods, including non-CS foods. Initial experiments demonstrated that conditioning a few exemplar food items through either method increased sensitivity to health and decreased sensitivity to taste when judging a variety of additional foods. Additional experiments using both methods replicated the generalization effect with regard to health sensitivity, but only when a task that preceded the EC procedure promoted, rather than interfered with, categorization of the CS foods by health. A push-pull experiment demonstrated that using specific food as CS produced effects comparable to those obtained when conditioning higher level categories, such as fruits and vegetables. We also demonstrated that the eating likelihood measure prospectively predicts healthiness of eating behavior outside of the lab. This research shows that EC can generalize to an entire dimension underlying food judgment and that this effect is facilitated by accessibility of the health dimension at the time of exposure to the EC pairings. |
