Take the 2-minute tour ×
User Experience Stack Exchange is a question and answer site for user experience researchers and experts. It's 100% free, no registration required.

When allowing users to provide feedback on other content, most websites allow users to provide both positive and negative feedback. StackExchange, for example, allows users to both upvote and downvote questions and answers, and YouTube allows users to both like and dislike videos.

Some websites, on the other hand, only allow users to provide positive feedback. Facebook is perhaps the most obvious example: users can "like" content, but not "dislike it". StackExchange kind of does this as well for comments: I can upvote your comment, but I can't really downvote it (though I can flag it, which is a bit different).

Are there any studies (either from actual user testing, or maybe from research in social psychology) on the effects of choosing one or the other? For example, why might Facebook not ever want to provide a dislike button? (I'm assuming there's a reason besides added software maintenance complexity.) Is it because adding a source of negative feedback might discourage some websites from adding like buttons (because while it's awesome to see that people like your content, it's pretty discouraging to see that even a single person doesn't), or is it perhaps because of reasons on the user side?

I'm looking for actual research on these questions, not armchair thought experiments.

share|improve this question
1  
You say that you want "actual research" on this, and by that I assume some sort of objective analysis. Can you think of a method for an investigation that could answer your question objectively without relying on "armchair analysis"? –  JohnGB Sep 24 '11 at 9:55
    
I am going to echo what JohnGB is implying here: finding objective studies about a very subjective and case specific question will be extremely hard to find. I am sure some of the larger sites like Facebook have done multivariate or A/B testing on such issues, but the results of these kinds of tests rarely get published externally. –  Nadine Schaeffer Sep 24 '11 at 19:18
    
I'd suggest changing the wording of your question to allow people here on StackExchange to at least provide you with their specific experiences with this kind of feedback. I find your question fascinating, but it is posed in a difficult manner. –  Nadine Schaeffer Sep 24 '11 at 19:20
    
I made the question intentionally difficult and objective so that it wouldn't be closed as being too "discussiony" and unanswerable. Unfortunately, I have no idea how I should word my question to avoid it getting closed. sigh –  grautur Sep 24 '11 at 19:33
1  
I think its a good question. And its quite likely that somebody's looked at the pre internet social psychology of say grading course papers. But if it happened pre 1997 - its not going to be searchable / or on free access. –  PhillipW Sep 24 '11 at 21:12
show 2 more comments

1 Answer

I'm not sure any research exists, but I have years of experience exploring and experimenting with this exact thing on my own. Since it may be the closest thing you can get in order to get an answer, my view:

Sites with upvotes only do not need their content to be accurate, they simply need content. For Facebook, "Like" is an interaction, and that is what they're after—any interaction IS their content, and the more subjective, the better.

Sites with both up and downvotes need the voting to mean something more universal, because they rank their content with it in a deeper way. StackExchange will die off if its voting system is flawed. If thousands of people "like" something not terribly interesting on Facebook, it doesn't injure Facebook the same way. Likewise StackExchange comments, since they are safely isolated from the primary product of the site (the value the community gives to the questions and answers) can be more of a social sandbox.

A Facebook "like" is always accurate because it's about what YOU like. A StackExchange "like" is part of an engine that exists primarily to discover the best of the best content in an objective sense. The latter needs the downvotes for accuracy, the former can be more casual about it and avoid the potential conflicts and hurt feelings(which could injure a social site) that negative voting brings.

Specifically, effects: Upvoting only keeps the community happy but doesn't get you accuracy, while adding downvoting will always make a permanent percentage of your users upset. (what this percentage is would be interesting to know, though it probably varies with the site's focus and people's willingness to take criticism in service to the larger cause)

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.