Slashdot Mirror

← Back to Stories (view on slashdot.org)

In 2012, Facebook Altered Content To Tweak Readers' Emotions

Posted by timothy on from the beats-the-top-stories-option-at-least dept.

The Atlantic reports that two years ago, Facebook briefly conducted an experiment on a subset of its users, altering the mix of content shown to them to emphasize content sorted by tone, negative or positive, and observe the results. From the Atlantic article: For one week in January 2012, data scientists skewed what almost 700,000 Facebook users saw when they logged into its service. Some people were shown content with a preponderance of happy and positive words; some were shown content analyzed as sadder than average. And when the week was over, these manipulated users were more likely to post either especially positive or negative words themselves. This tinkering was just revealed as part of a new study, published in the prestigious Proceedings of the National Academy of Sciences. Many previous studies have used Facebook data to examine “emotional contagion,” as this one did. This study is different because, while other studies have observed Facebook user data, this one set out to manipulate it. At least they showed their work.

7 of 130 comments (clear)

  1. consent by sribe · · Score: 5, Interesting

    There are laws governing obtaining informed consent from humans before performing psychological experiments on them. I doubt that a EULA can override them. This should be interesting...

    1. Re:consent by Sqr(twg) · · Score: 5, Insightful

      [citation needed]. Almost every major website does A/B_testing. Is there a law againt this? (That's not a rethorical question. I actually would like to know.)

    2. Re:consent by Ceriel+Nosforit · · Score: 5, Interesting

      There are laws against assault, bullying, and so on. The positive spin in innocuous but the negative spin is not.

      With 700 000 potential victims, the numbers are against them because when your sample size is that large outliers are the rule and not the exception.

      The risk of copycat suicide for example should have been obvious to those conducting this study.

      --
      All rites reversed 2010
  2. Ethical Responsibility by forand · · Score: 5, Insightful

    This is quite interesting research that should never have been done. I am rather surprised that the National Academy published the results of a study which violated multiple ethical guidelines put in place to protect human subjects. Did Facebook track the number of suicides in the 700,000 sample? Was the rate of those given a sadder than average stream have a higher or lower rate? Do the Facebook researchers address the ethical questions posed by performing such an experiment at all?

  3. As old Stanley Milgram would have said... by angularbanjo · · Score: 5, Funny

    ... that's pretty shocking.

  4. It's called the Common Rule by langelgjm · · Score: 5, Interesting

    It's called the Common Rule, although it generally only applies to federally funded research. There is some evidence that this study was in part federally funded. I think there are serious questions about whether a click-through agreement meets the standards of informed consent.

    Although the study was approved by an institutional review board, I'm surprised, and the comment from the Princeton editor makes me wonder how well they understood the research design (or how clearly it was explained to them). This would never have gotten past my IRB.

    --
    "Anyone who [rips a CD] is probably engaging in copyright infringement." - David O. Carson
  5. Future Facebook experiments! by QilessQi · · Score: 5, Interesting

    Hey, Facebook! Can you help me with some experiments of my own? I'd like to see the outcome if...

    1. Right before an election, all posts favoring candidate X or the political views of party X were promoted to the top of everyone's feed and given 20 extra fake "Likes", while posts favoring the opposition are demoted and de-liked.

    2. Phrases in posts favoring candidate X or the political views of party X are subtly "edited" when the appear in everyone else's news feed to be more positive (e.g., "like" to "love", "good" to "great"), while phrases in posts favoring the opposition are given the reverse treatment and sprinkled with misspellings.

    3. FB users with a tendency of opposition to party X have random fake posts/comments from them appear in other's feeds only, in which they insult their friends' baby pictures, make tasteless jokes, and vaguely threaten supporters of party X to "unfriend me if ur so lame u can't take the TRUTH, lol".

    --
    Koans and fables for the software engineer