On the last day of 2008, Harvard biologist John Bohannon gathered 18 guests in Brooklyn for some New Year’s Eve hors d'oeuvres. They had five spreads to snack on. Four of them were intended for human consumption and one was dog food. Bohannon wasn’t pulling a sick prank, but wanted to see if humans could tell dog food apart from the processed meaty products we normally eat. If they couldn’t, then we might just have a cheap substitute for that stuff.
The folks were offered duck liver mousse, pork liver pâté, liverwurst, Spam and dog food. All of the items were thrown into a food processor to give them a consistant consistency. Then, they were chilled, garnished, and placed out with Carr's crackers to see if people enjoyed their Scooby Snacks.
The ranking findings are not surprising. Humans do not like the taste of dog food. Over half the subjects, 72% (13 people) ranked the dog food dead last. The duck liver mousse was ranked first by 55% (10 people).
But here’s where things get weird. Although most subjects ranked Sample C, the dog food, as the worst tasting spread, when asked to report which sample was in fact dog food, most people did not guess Sample C! In fact, only 3 of the 18 subjects correctly identified Sample C as dog food! Instead, 44% (8 people) incorrectly chose liverwurst as the dog food.
As the authors suggest, the subjects could have been primed to expect the dog food to taste better than it did. After all, the dog food was Newman’s Own, and who doesn’t love Paul Newman? Additionally, since subjects were assured that their “experience would not be disgusting, they might have excluded the worst-tasting sample from their guesses.” [Dog Spies]