Why can calorie posting be apparently ineffective? The roles of two conflicting learning effects

We investigate why the aggregate effect of calorie posting on calorie consumption can be insignificant by decomposing the learning effect into two conflicting components: a calorie-decreasing effect of learning that one was underestimating caloric content (LUE), and a calorie-increasing effect of learning that one was overestimating caloric content (LOE). Our lab snack-order experiments demonstrate the existence of the LUE effect (−8.3%) and the LOE effect (+4.8%), where the aggregate learning effect is −5.8%. Our results also imply that the LUE can be cancelled out by a positive saliency effect, while the undesirable saliency effect may be mitigated by combining the calorie posting with information about daily calorie needs.