Sunday, January 4, 2009

The Interactive Effect of Cultural Symbols and Human Values on Taste Evaluation

ResearchBlogging.orgI've previously reported on a few studies showing some of the ways that taste evaluation isn't a simple response to the physical properties of what is tasted. Actually, now that I check I've only blagged about one so far: We've seen how the same wine (at least when delivered as sips that aren't being paid for) is preferred when it is thought to cost more. But there will be more on this topic, because I enjoy running interference at my wine tasting group.

This paper reports an attempt to find out whether a sense of congruence with an individual's values makes a difference to how good something tastes to that individual. The answer turns out to be yes. Here's the abstract:
We suggest that consumers assess the taste of a food or beverage by comparing the human values symbolized by the product to their human value priorities. When there is value-symbol congruency, they experience a better taste and aroma and develop a more favorable attitude and behavior intention; incongruence has the opposite effect. Participants in two taste tests were told the correct identity of a product or misinformed. Participants who endorsed the values symbolized by the product (that they thought they were tasting) evaluated the product more favorably. The implications for marketing strategy, self-congruity theory, and the assimilation effect are discussed.
This certainly isn't a crazy hypothesis. As the authors of this study note, we've known for ages that people rate the very same carbonates water more highly when they're told it's Perrier (Nevid 1981). Also, the effects aren't always consistent, because different people have different expectations - differing for example over whether 'healthy' or 'organic' food will taste better or worse.

To test the hypothesis in this case two different taste tests were set up, one for a food and one for a drink.

In the food case omnivorous participants tasted a vegetarian meat substitute (chosen to be least distinguishable from actual meat) sometimes thinking that it was meat, and sometimes that it was substitute. It had been previously established that in the relevant population the vegetarian versus meat eating distinction was associated with low versus high preference for 'social power'.

In the drink case two colas, one a brand Cola (Pepsi) found to have high association with an 'exciting life' and another store branded (in this case Australian Woolworth's) were offered. The colas had been previously found to be indistinguishable in blind tests, and subjects again sometimes drank falsely believing they were drinking another cola.

All participants completed a value questionnaire (to assess their values with reference to social power, and exciting lives) and a multi-scale taste assessment of the products they had tasted. The scales were combined to give a single linear taste assessment for each subject. Participants were randomly assigned to taste what they were told they were tasting, or to taste under a misconception. They were also asked about attitude and purchase intention, current food and drink consumption, social desirability bias, and asked whether they believed they had been drinking or eating what they had been told.

First, it was found that:

In the drink category participants irrespective of condition preferred Pepsi to the Woolworth's. (In the actual experiment they had a cup of it. It's not clear whether the earlier test that found them indistinguishable had been a sip test.)

In the food test participants showed no main effect of actual product in their taste evaluations.

More significantly, the main predictions were confirmed: "low social power participants gave the food a higher taste evaluation when they believed they had tasted a vegetarian alternative and gave the food a lower rating when they believed they had tasted a sausage roll (b = .42, t = 4.1, p < .001). High social power participants showed the opposite tendency, though to a lesser degree (b = .11, t = .9, p = NS)."

And: "participants who valued excitement and enjoying life had a more favorable attitude and purchase intention when they believed they had tasted Pepsi than when they believed they had tasted Woolworth (b = .44, t = 4.5, p < .001)."

These results are analyses factoring in various elements of the different questionnaires - you should read the paper for the details, which would slow down the exposition here. This is further evidence that taste evaluation is not a simple or direct response to the physical properties of the thing tasted, but depends in significant ways on the symbolic and other associations of the thing tasted. I reckon you could find similar effects for foods presented as being 'local' or 'foreign' for people with strongly nationalist or xenophobic tendencies.


Nevid, Jeffrey S. (1981), Effects of Brand Labeling on Ratings of Product Quality, Perceptual and Motor Skills, 53, 407-10.

Michael W. Allen, Richa Gupta, Arnaud Monnier (2008). The Interactive Effect of Cultural Symbols and Human Values on Taste Evaluation Journal of Consumer Research, 35 (2), 294-308 DOI: 10.1086/590319


Michael Meadon said...

Interesting stuff...

And shouldn't that be wine 'tasting'?

Doctor Spurt said...

Dude, it is 'tasting' on my current browser. I'll check from the office...