Certainty & Uncertainty in Restaurant Reviews

[Chris Amirault]

In "Defend Your Research: Experts Are More Persuasive When They’re Less Certain," just published on the Harvard Business Review website, Zakary Tormala writes compellingly about persuasion and certainty -- and he uses restaurant reviews as the basis for the article.

[Tormala] had subjects read a review that gave a new eatery four out of five stars. Half the subjects read a write-up by a professional food critic, and the other half, one by an amateur. Half of each group saw a review that was highly certain; the rest, a review that was tentative. The subjects were then asked how good they thought the restaurant would be. While the confident amateur inspired subjects to give better ratings than the uncertain one, the less assured expert prompted higher ratings than the certain expert.

In response to a skeptical interviewer, Tormala pointed out,

[Y]ou have to separate confidence in facts from confidence in opinions. Research has shown that in court testimony, confidence makes an eyewitness more credible, that it can impact trial outcomes. But there is a right or wrong answer there. In our research, we’re dealing with tastes and attitudes.

Well, we deal with tastes and attitudes around here quite a bit, so I thought I'd put the question to you. Me? I'm still uncertain....

[Holly Moore]

Pretty much explains the success of Yelp and Zagat.

[Chris Amirault]

I think that Yelp and Zagat have droves of quite certain people posting their opinions.

[runwestierun]

Maybe people associate uncertainty with humility and trust that, and maybe certainty smacks a little of pontification.

[MaxH]

"People expect experts to be confident," says Prof. Tormala, but more is going on in restaurant criticism than the linked interview ever mentions. Readers know that critics draw a picture from limited experience of a restaurant. Readers also want useful information (helping to predict their own reactions there). It's not clear to me, from the interview, that the uncertain pro reviews had more impact because of the rather general "interpersonal dynamics" of "expectancy violations" that interest Tormala, or because their tentativeness cued readers that they were reading a careful reporter.

-- Because "tastes and attitudes" notwithstanding, good restaurant critics also report a lot of facts and observations, and give basis for conclusions. That may be humble but it's also writerly prudence, not pretending omniscience or overreaching your real data.

Contrast the hokiest restaurant critiques you've seen (and that some Web sites seem almost to encourage). Conclusions without basis; wide generalizations from a single (sometimes no) experience of the restaurant. Little uncertainty there. But I'm considering only the "pro" review results, and it's hard to understand any of the results deeply without knowing the literal texts that the subjects saw.

[Jon Savage]

I think that Yelp and Zagat have droves of quite certain people posting their opinions.

The thing with yelp and Zagat is there are many clueless people confidently posting their opinions. On example in point was a beer place that just opened nearby that has a pretty spectacular selection of well kept and served draft beers. I think it was averaging am scant 2 stars on yelp.

The complaints?

1. The beer(lambic)was sour and tasted funnny (that is how they are supposed to taste??)

3. The pours of Belgian doubles, triples etc. were too small (you don't drink this stuff by the pint??)

4. many of the beers were served in odd shaped e.g. non pint glasses (you mean there are different shaped glasses for different kinds of beer??)

...I could go on and on but suffice to say this place is doing all the right things and serving some spectaular beers very well indeed. Without really reading those reviews I might never have gone there at all based on the rating.

[MaxH]

...a beer place that just opened nearby that has a pretty spectacular selection of well kept and served draft beers. I think it was averaging am scant 2 stars on yelp. ... suffice to say this place is doing all the right things and serving some spectaular beers very well indeed.

Those situations (I've seen plenty) are actually counterproductive, perverse, for the reader-consumer. They might make money for the Website owner (which has tried to sell advertising whenever restaurarateurs I know complained of misleading comments there), but a reader actually has a clearer picture of the establishment _without_ seeing the "stars" or even some of the comments,

Such sites seem to draw out the least useful commentators. With chips on their shoulders, or personality disorders, or who actually, seriously, judge a whole restaurant on some stupid little detail like those 6ppc quoted (I've seen plenty of that). Their comments end up communicating about themselves and nothing, or less than nothing, about the restaurant.

[nvgoodtimes]

I use Yelp a lot to get some feed back on a restaurant. You have to take each review with a grain of salt as there is little consistency in each person's rating system. Typically you will find some reviews at 2 stars and some at 5 stars. However if all reviews are low/bad it does ring the warning bells. You can have a bad experience even at an excellent restaurant.

Another problem is that many people will only post a bad review. They never get around to writing about their positive experience. I tend to also at other reviews posted by the reviewer to see if they are posting fair reviews or tend to be biased on way or the other. Still, any information is better than none. If you do not agree with the reviews posted, you can write your own.

[heidih]

If one draws on the general public then certain criteria and expectations apply. For myself, I am more interested in the description of the food as opposed to the conclusion. Some are well written and give you all the info. Some just state their final "decision" - those I ignore.

[Edward J]

From another point of view.....

Once you do achieve a 4 or 5 star rating in uh.."certain" websites, it's time for someone to cash in. Either it's the website looking for advertising, blooggers looking for freebies, or "customers" who, by their descriptions, make it clear they have never been in the place, make vague complaints with absoluelty no descriptions to back them up in order to drag the point sytem down.

DAMHIKT...................

[Mjx]

I've been offered money to write reviews. As far as I'm concerned, that's enough to undermine faith in any guide.

[gfweb]

An unknown reviewer is useless to me. I need a track record of accurate reviews before I pay attention. The Yelp/Zagat approach is just pointless. People give favorable reviews to what they like, which could be slop. There are people who actually like a McDonald's burger.

[tino27]

An unknown reviewer is useless to me. I need a track record of accurate reviews before I pay attention. The Yelp/Zagat approach is just pointless. People give favorable reviews to what they like, which could be slop. There are people who actually like a McDonald's burger.

This drove me nuts, too. Prior to Yelp's popularity and me starting my own food blog, I use to leave mini-reviews on Yahoo. It would drive me nuts to see people rate a McDonald's with five (out of five) stars. It made me realize just how worthless a "star" rating can be. I also realized fairly quickly that anything I had to say about the food or my experience would get lost to all of the other posters who don't seem to understand that USING CAPITALS LETTERS FOR THE ENTIRE REVIEW AMOUNTS TO SHOUTING.

To address the original purpose of this thread, I personally try to write restaurant reviews with as many facts as possible, explain any opinions that I state, and let the reader watch my thought process so that they can see themselves sitting at the table with me (or avoiding it if the food/service was bad). I think it requires a certain amount of humility as well, as you have to remember that you are commenting on people's livelihood, not just food in the abstract.

[MaxH]

I have to assume that the people who claim to rely on Yelp (which often surfaces when this subject's discussed online, in my experience) simply have not yet noticed the side of it that I and 6ppc have identified as harmful. The side that demonstrates concretely, to almost anyone, that "some" information is not always better than no information, if it's badly misleading.

You notice this mainly when examining comments about a restaurant that you know extremely well (i.e., better than the commentators) so that you are actually appraising the commentators, against a known reference point. That was my whole introduction to Yelp, years ago, when people pointed out that an independent local restaurant, which we all knew intimately, was accumulating wildly misleading and unrepresentative comments. Content contrary to the implied purpose of restaurant criticism to usefully inform and advise. Unfortunately this hasn't changed, I continue to see the same thing when I track restaurants that I know well enough to be able to judge their commentators.

ETA:Last night a restaurateur said, unsolicited, that he'd been offered the privilege to "edit" reviews of his restaurant for $1000 a month. That's the latest of many pay-to-play reports I've heard for years from restaurateurs. I guess this might help with the misleading-comments problem, but it introduces another one.

Edited February 23, 2011 by MaxH (log)

[Dakki]

Maybe people associate uncertainty with humility and trust that, and maybe certainty smacks a little of pontification.

I think there might be something to this, or a related effect: I have an idea that people who are knowledgeable on a given topic tend to go on and on about inherent uncertainties and special cases while the guy who thinks he heard something in a bar once will state that bald "fact" as absolute truth.

No opinion on Yelp. Never used it.