Experts Can Be More Persuasive if They Express Uncertainty According to Stanford Business School Research

STANFORD, Calif.--(BUSINESS WIRE)--Zakary L. Tormala’s research flies in the face of logic. If you’re an expert and make your points with confidence, people will be far more convinced than if you sound uncertain. Right?

“Bianco’s — a tentative 4 out of 5”

Well, no — at least not when it comes to consumers, the associate professor of marketing discovered in research he coauthored with Stanford Graduate School of Business doctoral candidate Uma R. Karmarkar and reported in today’s StanfordKnowledgebase.

“Our key finding,” Tormala said, “is that although non-experts can become more persuasive by expressing high certainty about their opinions, experts can become more persuasive when they express some degree of uncertainty. Across several studies, we found that expert sources gained interest and influence by expressing minor doubts about their own opinion.”

Tormala said incongruity between the source’s expertise and level of certainty makes his or her message more intriguing. “Whether it’s a person without established expertise in a given domain expressing very high certainty, or a person with clearly established expertise in a domain expressing low certainty,” Tormala said, “the inconsistency is surprising. It draws people in. And as long as the arguments in a message are reasonably strong, being drawn in leads to more persuasion.”

Earlier research by others had made the case that expressing certainty generally increases people’s persuasive power, because it boosts their perceived credibility. Tormala pointed out that those studies concerned topics such as witnesses testifying in court or stock market advisers giving stock recommendations where there is an objective truth or correct answer. In those instances, he said, people might rely on a person’s certainty as an indicator of his or her credibility. “In more subjective domains like consumer contexts, though,” Tormala said, “expressing certainty appears to have a more dynamic effect, giving a message more or less impact depending on who is expressing it.”

Tormala and Karmarkar studied this issue by giving consumers what purported to be customer reviews for a new restaurant. “Restaurant reviews provided a good setting for this research,” Tormala said. “Like other consumer topics, they’re subjective, but there also are traditional markers of quality.”

He said such attributes as ambience, service, and taste of the food can be described with enough detail to let people understand the reviewer’s perspective, but still reach their own conclusions about whether the restaurant would be suited to their tastes.

Participants in one experiment read a favorable review of a new restaurant called “Bianco’s.” Across experimental conditions, the main part of the review contained the same core comments about Bianco’s, comments the researchers had pre-tested on a sampling of readers to make sure they were strong. In the main study, some participants were told that the reviewer was a renowned food critic who often contributed to a major regional newspaper; others were told that the reviewer was a network administrator at a local community college who kept a personal web journal — and normally ate fast food.

In addition to varying the supposed source of the review, the researchers varied the level of certainty expressed in it. In the high-certainty recommendation, the review was titled “Bianco’s — a confident 4 out of 5,” and the author expressed certainty about the quality of the food and the restaurant twice in the review (saying, for example, “Having eaten there for dinner, I can confidently give Bianco’s a rating of 4 [out of 5] stars”). In the low-certainty recommendation, the title of the review was “Bianco’s — a tentative 4 out of 5” and the author expressed uncertainty about these same points (for example, “Having eaten there only once, I don’t have complete confidence in my opinion, but I suppose would give Bianco’s a rating of 4 [out of 5] stars”).

“We find that when the regular, everyday person is extremely certain, that’s surprising to readers,” Tormala said. “Conversely, when the expert is not so certain, that’s surprising. In both cases, surprise increases readers’ interest in and involvement with the review, which is essentially a persuasive message, and this promotes persuasion. So non-experts get more attention and can have more impact when they express certainty in their messages. Experts, in contrast, get more attention and can have more impact when they express uncertainty.”

Tormala added that all the findings apply only if the basic message is compelling, or contains strong arguments. When a fictional reviewer wrote about liking the restaurant because he approved of the colors on the menus or enjoyed a conversation with a friend, his opinions had little impact on readers regardless of how much certainty he expressed.

“This ties our experiments to a classic finding in persuasion research — increasing consumers’ involvement and processing of your message is a good thing as long as your message is strong. If your message is weak, increasing others’ involvement or interest has no effect, or can even backfire.”

(This story reports on research at the Stanford Graduate School of Business and appears in today’s Stanford Knowledgebase, the free monthly information source for thoughts, ideas and research at the Stanford Graduate School of Business. To dig deeper, visit: http://www.gsb.stanford.edu/news/research/tormala_certainty.html?cmpid=kbpage&edition=09-oct.)