TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Why the Facebook Experiment Is Lousy Social Science

122 点作者 dnt404-1超过 10 年前

3 条评论

gwern超过 10 年前
I wrote a comment debunking OP&#x27;s arguments, but it was too long for a HN comment and my noprocast settings kicked in so I can&#x27;t post it as two comments; so I copied it over to G+: <a href="https://plus.google.com/103530621949492999968/posts/1PqPdLyzXhn" rel="nofollow">https:&#x2F;&#x2F;plus.google.com&#x2F;103530621949492999968&#x2F;posts&#x2F;1PqPdLyz...</a><p>tl;dr: Let&#x27;s summarize his complaints and my counter-objections:<p>1. no consent: irrelevant to whether this was good science or &#x27;lousy social science&#x27; 2. crossed boundaries between corporations and academia: likewise irrelevant; also, welcome to the modern Internet 3. small effect size: misunderstood the statistical design of study and why it was designed &amp; expected to have small effects 4. used LIWC with high error rate for measuring emotionality of posts: if random error, biases effect to zero and so is not an argument against statistically-significant findings 5. and LIWC may have systematic error towards positivity: apparently not an issue as negative &amp; positive conditions agreed, and the studies he cites in support of this claim are mixed or unavailable 6. also, other methods are better than LIWC: sure. But that doesn&#x27;t mean the results are wrong 7. maybe LIWC has large unknown biases applied to short social media texts: possible, but it&#x27;s not like you have any real evidence for that claim 8. Facebook news posts are a biased source of mood anyway: maybe, but they still changed after random manipulation 9. experience sampling is sooooooo awesome: and also brings up its own issues of biases and I don&#x27;t see how this would render the Facebook study useless anyway even if we granted it (like complaints #1, 2, 6, 7)<p>Now, I don&#x27;t want to overstate my criticisms here. The author has failed to show the Facebook study is worthless (I&#x27;d wager much more money on the Facebook results replicating than 95% of the social science research I&#x27;ve read) and it would be outright harmful for Facebook to aim for large effect sizes in future studies, but he does at least raise some good points about improving the followup work: Facebook certainly should be providing some of its cutting-edge deep networks for sentiment analysis for research like this after validating them if it wants to get more reliable results, and it would be worthwhile to run experience sampling approaches to see what happens there, in addition to easier website tests (in addition, not instead of).
评论 #8380685 未加载
评论 #8381244 未加载
kylelibra超过 10 年前
The uproar over this particular experiment seems like a bit much, but it doesn&#x27;t hurt to start to have these conversations about research on users using the web. There is only going to be more of it going forward.
评论 #8380363 未加载
gpanger超过 10 年前
Hi guys, author here, thanks for taking an interest in my critique of the FB experiment. Quick FYI is that the post is up on Medium:<p><a href="https://medium.com/@gpanger/why-the-facebook-experiment-is-lousy-social-science-8083cbef3aee" rel="nofollow">https:&#x2F;&#x2F;medium.com&#x2F;@gpanger&#x2F;why-the-facebook-experiment-is-l...</a><p>...where I can fix things like broken links. It was republished on my school&#x27;s website, but unfortunately I don&#x27;t have control over the HTML there. In the cases where stuff broke, it&#x27;s because I linked to the author&#x27;s personal manuscript and not the official journal page (because the former is freely available to everyone rather than behind a paywall).<p>I care a lot about Big Data research, especially involving social media, and think we too often ignore the conceptual leaps required to make inferences about the human experience from social media.<p>Here, the leap is: sentiment_analysis(what people say on social media) == how they really feel.<p>The point about LIWC (the sentiment analysis tool used here) is that (a) it&#x27;s flawed, and perhaps not in a &quot;random error&quot; kind of way, (b) we don&#x27;t really know how well it works because it has not been validated in social media or Facebook posts specifically, which should make most researchers nervous (but somehow doesn&#x27;t), and (c) there&#x27;s evidence from other data sources that suggests LIWC is biased in a way that would underrepresent the emotions of interest (namely low arousal emotions like sadness, depression, loneliness, feeling left out, etc.; these aren&#x27;t picked up by LIWC as well as other negative emotions like anxiety).<p>See e.g.:<p>O’Carroll Bantum, E., &amp; Owen, J. (2009). Evaluating the Validity of Computerized Content Analysis Programs for Identification of Emotional Expression in Cancer Narratives. Psychological Assessment, 21, 79-88.<p>The point isn&#x27;t that using LIWC means the experiment is invalid, the point is that it should give us pause and caution us against stating the conclusions of the experiment too strongly. I think the authors do state their conclusions a bit strongly.<p>The other main critique is about biases inherent in social media as a datasource itself. The private, randomly-solicited emotion samples of experience sampling are more likely to capture Facebook&#x27;s true emotional impact than the non-private, self-selected emotion samples of status updates. Let&#x27;s just take arousal bias. If we know that people are more likely to post when they&#x27;re emotionally aroused (excited, angry, fearful&#x2F;anxious), but the emotional consequences we&#x27;re concerned about involve low arousal emotions (sadness, depression), then there&#x27;s a serious chance we&#x27;ll miss exactly the emotions we&#x27;re arguing don&#x27;t exist. That&#x27;s a big problem.<p>I think we&#x27;re a bit too enamored with the idea that Big Data provides an unbiased window into the human experience. I think a tremendous amount of social science would argue otherwise.<p>Stepping back a bit, the Facebook experiment raised many interwoven issues for me, which is why they featured in the broader piece I wrote. About Facebook&#x27;s culture, about ethics in corporate vs. academic research, about Facebook&#x27;s emotional impact (foolish to believe there is none, I think), about how we use Big Data in research, about how we cope with Facebook&#x27;s presence in our lives, for better or worse.<p>You may have a great experience with Facebook, and that&#x27;s great. Others struggle with the medium. I mention social comparison not just because it&#x27;s been the focus of research on social media, but also because the authors of the experiment bring up social comparison (as well as the &quot;alone together&quot; argument) in their work. Because they try strenuously to rebut the unfavorable findings about Facebook&#x27;s emotional consequences, I thought it was important to point out that their study seems designed in a way that would systematically underrepresent exactly those negative emotions they&#x27;re arguing against.<p>Certainly, some negative emotions were &quot;contagious&quot; through social media (anxious news reports, for example), as were some positive emotions. But is the emotion that gets retransmitted the full emotional picture? Probably not. Probably many emotions and feelings get withheld. The social science would suggest that when positive posts make us feel bad, we won&#x27;t go back on Facebook and broadcast those feelings to all of our friends.<p>Thanks a ton for engaging with my critique and responding with your own. Writing that was a labor of love, and I learn a lot from the feedback, good and bad. Sorry it was so long.