Prior Experience Shapes How Consumers Compute New Information
Over time, consumers develop a set of cues that we then use to make inferences about products, such as “all French restaurants have great service” or “more expensive candles smell better.” However, this set of predictable beliefs can make it difficult for us to learn and recognize other real, positive qualities that are indicated by the same cues, reveals a new study.
“Once people learned that a cue predicted an outcome, they became less likely to learn about this very same cue with respect to a different outcome,” write Marcus Cunha Jr. (University of Washington), Chris Janiszewski, and Juliano Laran (both University of Florida). “The implication is that the learning system is designed to discourage single cue–multiple outcome learning.”
In the pilot study of a series of five experiments, the researchers used cheese tasting to explore the development of predictive knowledge structures, a phenomenon also known as “protection of prior learning.” They first had participants taste an orange rind Raclette cheese that was mild and creamy, and a purple rind Drunken Goat cheese that was much stronger tasting and dry. They then had participants rate the cheeses on a scale of mild to strong to induce the association with an orange rind and a mild flavored cheese. A control group also tasted two different types of cheese but did not rate them.
To test whether an association between an orange rind and mild flavor would make it more difficult for consumers to gauge other existing qualities, such as texture, tasters were then asked to rate the creaminess of a mild, creamy Port Salut with an orange rind and a dry Manchego with no rind. Surprisingly, participants were less likely than the control group to expect the second orange rind cheese to be creamy, even though the first one had also been creamy. As the researchers explain, “Learning that the orange rind predicted a difference in the strength of flavor . . . attenuated the learning that the orange rind predicted creaminess.”
This research has important implications for marketers, policy makers and consumers. For instance, the researchers point to Merck’s introduction of the cholesterol-lowering drug Simvastatin under the brand name Zocor. Recently, researchers found that Simvastatin may be also effective at preventing the onset of Alzheimer’s disease.
“This opportunity creates a branding dilemma for Merck,” the researchers write. “Our findings suggest that consumers may be slower to learn the Alzheimer’s relief association to [Zocor] than to a new brand name.”
Similarly, from a public policy standpoint, the results suggest that people may be resistant to adopt new health and safety standards when information conflicts with prior learning. Beyond creating awareness, successful campaigns might present new information in a way that does not utilize attributes already associated with another outcome.
Journal of Consumer Research, Page 000–000, DOI: 10.1086/523293 Electronically published October 10, 2007