Effects of sampling healthy versus unhealthy foods on subsequent food purchases

Food sampling at retail stores and restaurants (e.g., amuse bouche) is a widespread practice. These food samples vary considerably in healthfulness levels. Prior research has primarily focused on the effects of sampling on evaluations and sales of the sampled item. However, can there be unintended consequences of sampling a healthy versus an unhealthy item on subsequent purchases of other food items? Also, would the degree of dissimilarity between the sampled item and subsequent options moderate the effects? The results from a series of experiments, including four studies conducted in field settings, show that sampling a healthy (vs. unhealthy) item paradoxically leads to greater subsequent purchase/choice of unhealthy foods – but only when consumers perceive a relatively high level of dissimilarity between the sampled item and subsequent options. This effect reverses when the sampled food and subsequent options are perceived as being relatively low on dissimilarity (i.e., high on similarity).