Consumers' ratings of the natural and unnatural qualities of foods.

An investigation sought to understand what consumers perceive by the term natural. The aim was to test eight hypotheses on food ingredients and processes used for manufactured food. A representative sample (n=190, aged 18-65 years), rated 50 food exemplars for naturalness (0-100 scale). Data were analysed by repeated measures ANOVA. Results support three hypotheses: chemical changes were more potent than physical changes; there was a minimal effect of mixing like entities and the more processing the greater the effect on consumer's deviation away from natural. Two hypotheses were validated conditionally: contagion accounts for naturalness reduction but is independent of dose above a certain level; E-numbers were always perceived to be less natural than the same preservatives described by chemical and common names; however, there were gender and some education interaction effects. The hypothesis that addition has a greater effect than removal was only partially validated. There was no evidence found to support the hypotheses that process has more effect than content, or that novel ingredients have a greater effect than 'known' ingredients, however, this result may have been confounded. The implications for new manufactured food products, suggested by the results, are that products with physical changes, less processing, with like ingredients and described using common named descriptors for ingredients would be perceived to be more natural. Copyright 2010 Elsevier Ltd. All rights reserved.