Slashdot Mirror
Why You Shouldn't Trust Internet Comments
sciencehabit writes "A new study suggests that all the reviews you read on Yelp and Amazon are easily manipulated. It's not that companies are stacking the deck, necessarily, it's that a few positive comments early on can influence future commenters. In fact, when researchers gamed the system on a real news aggregation site, the items received fake positive votes from the researchers were 32% more likely to receive more positive votes compared with a control (abstract). And those comments were no more likely than the control to be down-voted by the next viewer to see them. By the end of the study, positively manipulated comments got an overall boost of about 25%. However, the same did not hold true for negative manipulation. The ratings of comments that got a fake down vote were usually negated by an up vote by the next user to see them."
Nothing new here, move on...
This is by far the most insightful treatment I've ever read on this important issue. Everyone who does business on the Internet must read this valuable study.
My hypothesis:
1) Products with positive comments are more likely to be purchased.
2) People identify themselves by their choices, and no one wants to make a bad choice. Ergo, almost by definition, any choice people make is "the right one". At the very least, people are predisposed to liking what they spent money on.
Remember; When discussing all things retail, it's not how good the product is, but how well it satisfied the need. At least half the time, that need is largely imaginary.
Sure, I could RTFA, but this is more fun.
Mod me down with all of your hatred and your journey towards the dark side will be complete!
I'm always glad to hear about research like this myself, but this has severe ethics problems. You don't con people to show how easy it is to con the people. I know that rationalization is popular with some segment of you "hackers" out there, but whenever social scientists do this, they end up getting hasled about it.
I'd say it's far more complicated then that.
If you have a bad experience and go to the product review and the other reviews are bad, you are apt to write a review confirming what you are reading.
On the other hand if you have a bad experience and all the other product reviews are good you may have a moment of self doubt (did I mess up with the product) which makes you less willing to post a negative review.
When I check reviews, one of the first things I do is check the negative reviews. Why? Because half the one-stars are often jackasses with no clue what product and/or service they were buying. Other times, knowledgable and otherwise reasonable people have found the service or product being rated to be inadequate in some significant way.
And then I look for high ratings to see if they are reviewing the product in a reasonable manner. From there, I make my own decisions regarding the validity of both sides.
Anyone who decides just based on the stars/review-based numbers is a fool.
In SOVIET RUSSIA... erm...NSA AMERICA, the Internet logs onto YOU!
just imagine a beowulf cluster of insensitive clods!
have you seen my sig? there are many others like it but none that are the same
I hate to say this, but I find I am more likely to take the time to write a bad review than a good one. (Anger is a great motivator.) I assume others are like this as well so I read the negative reviews in that light.
Also, any review in all caps, good or bad, I automatically discard.
But how much less likely are you to down-mod a score-5 tweet than a score-1? And how much more likely are you to read-and-upvote a red firehose submission than an indigo?
I hate to say this, but I find I am more likely to take the time to write a bad review than a good one. (Anger is a great motivator.) I assume others are like this as well so I read the negative reviews in that light.
Also, any review in all caps, good or bad, I automatically discard.
That's been my experience, too. Anger motivates you to want to do something, so people lash out on the comment board. People who are satisfied, by definition, aren't really motivated to take any additional steps.
Today's Sesame Street was brought to you by the number e.
I'll leave positive comments, normally when I receive outstanding customer service. People should be recognized when they do good work, and far too often management doesn't bother unless they have input from outside. I'll also ask to speak to their supervisor if the situation warrants it, and let them know if someone has done something outstanding.
"Think about how stupid the average person is. Now, realise that half of them are dumber than that." - George Carlin
Or maybe they just weren't visible to the moderators until a few upvotes brought them above the viewing threshold. Once visible, comments that happen to be genuinely insightful, informative etc, would get more upvotes because they deserve it. You don't have to invoke bandwagon effects to describe what you observe. Also, when I have moderator points, I tend to upvote good comments that *don't* have a high score because they are worth drawing attention to. Wasting moderator points on a "me too" upvote of a +5 comment is a poor use of the privilege. In my view, the purpose of moderation isn't to "skew" the discussion to reinforce the echo chamber., Rather moderation should improve the overall signal/noise ratio so threshold settings are actually meaningful.
Everytime I look to buy something I am not not an expert on or are on the fence about quality, I look up reviews and sort by rating. And I look at negative ratings first because a vendor or retailer won't pad a product with negative reviews. And even though there will always be negative reviews from people who dont have a clue or give little to no info, there are som rea gems out there that give you a clear picture of what you are getting into.
Obligatory anecdotes:
When looking on newegg I sort by lowest score first and read the reviews. You always have some dummies who obviously have no idea what they are doing and rate 1 star because of a mistake they made. But you also run across some genuinely informative negative reviews which are more influential to me than positive reviews. For example, I was looking to buy a uATX board from ECS that had the AMD bobcat CPU onboard. It was perfect, had extra PCI shots for SATA cards for a low power Linux server box. It turns out in a few of the negative reviews there was an IRQ bug that severely impacted performance. I was close to buying it but then scratched it off my list.
Recently I was also looking to purchase a generator from a coworker who bought it after hurricane sandy but never used it, the box is unopened. The brand name was Generac and I have owned two other Generac products, a power washer and a 4kw generator. Both of those machines went south after little use, the generators exhaust valve stuck open when the valve seal went bad and allowed oil to seep down the valve and seize it (I fixed that but it never ran quite right, stalled and was a bitch to start). The power washer engine needed its carb rebuilt and then the water pump blew a shaft seal. But that was 6+ years ago and I figured Generac got their shit together by now. After reading negative reviews on Amazon I came to understand that Generac will try to weasel out of warranty repairs and "authorized" repair shops frequently change as they get shafted after Generac refuses to reimburse them for warranty repairs already performed. There were also negative reviews that warned of blown stator coils after a few hours use and lemons that wouldn't start out of the box. Since the generator could not be returned to the original vendor I took a pass even though I would get a sweet deal ($200 off retail as he wanted to dump it). Maybe it would have worked fine but I didn't want to risk losing 800 bucks and damage my business reputation.