Can People Distinguish Pate from Dog Food? Apparently Not.

by robekulick

A recent working paper for the Journal of Wine Economics asked the age old question, can people reliably distinguish pate from dog food? Well maybe its not an age-old question, in fact, until I saw the paper I had never made the connection myself, but nevertheless three enterprising researchers asked and answered this question in a double blind taste test:

http://www.wine-economics.org/workingpapers/AAWE_WP36.pdf

The researchers presented subjects with five unlabeled blended meat products. One of the meats was Newman’s Own dog food, which was placed in a blender to give in the appearance and texture of liver mousse. The other meat products in the test were duck liver mousse, pork liver pate, supermarket liverwurst, and Spam.  Each of these products was also put in a blender so that it had the texture and appearance of liver mousse.

Subjects were first asked to rank the meats in terms of taste. They were then asked to identify which of the meats was the dog food.

The researchers found:

“The dog food (sample C) was ranked lowest of the five samples by 72% (13) of subjects.  The duck liver mousse (sample A) was rated as the best by 55% (10) of subjects.  Between these extremes, the majority of subjects ranked Spam, pork liver pâté, and liverwurst in the range of 2nd to 4th place (see Table I and II).

The aggregate taste ranking of the dog food was highly significant (see Table III).  The ranking difference between dog food and Spam was greater than the P<0.05  threshold, and the difference was greater than the P<0.01 threshold for all other samples.  Subjects’ preference for the duck liver mousse was also highly significant. The only  sample that was not ranked significantly differently than the duck liver mousse (at the  P<0.05 level) was the pork liver pâté.”

In other words, this part of the test went as you might expect. People’s preference for the duck liver mousse was statistically significance and the ranking of the dog food as last was also statistically significant.

But then, something strange happened. The subjects were not able to systematically guess which food was the dog food.

Now the paper had a small sample and I’m sure it wasn’t really a random cross-section of the population so it may not be representative, but its still pretty interesting. What’s clear is that after deciding what food tasted the worst people did not automatically think the worst tasting food was the dog food. In other words, people apparently assumed that some human foods taste worse than premium dog foods.

However, if the comparison had just been between the duck liver mousse and the dog food, I suspect the outcome may have been very different. The result may just reflect that people don’t find liverwurst and Spam particularly appealing.

 

Advertisements