Evaluating source credibility effects in health labelling using vending machines in a hospital setting.

Objectives: Providing advice to consumers in the form of labelling may mitigate the increased availability and low cost of foods that contribute to the obesity problem. Our objective was to test whether making the source of the health advice on the label more credible makes labelling more effective.

Methods And Measures: Vending machines in different locations were stocked with healthy and unhealthy products in a hospital. Healthy products were randomly assigned to one of three conditions (i) a control condition in which no labelling was present (ii) a low source credibility label, "Lighter choices", and (iii) a high source credibility label that included the UK National Health Service (NHS) logo and name, "NHS lighter choices". Unhealthy products received no labelling. The outcome measure was sales volume.

Results: There were no main effects of labelling. However, there were significant interactions between labelling, vending machine location and payment type. For one location and payment type, sales of products increased in the high credibility label condition compared to control, particularly for unhealthy products, contrary to expectations.

Conclusions: Our findings suggest that high source credibility health labels (NHS endorsement) on food either have little effect, or worse, can "backfire" and lead to effects opposite to those intended. The primary limitations are the limited range of source credibility labels and the scale of the study.