[Previous] People are Wrong then Ignore Criticism | Home | [Next] No Path Forward

Bad Scholarship by Wall Street Journal

The Wall Street Journal (WSJ) doesn't care about scholarship like accurate quoting. They write:
According to an abstract of the study, "for people who had positive content reduced in their News Feed, a larger percentage of words in people's status updates were negative and a smaller percentage were positive. When negativity was reduced, the opposite pattern occurred."
They link the study and the abstract doesn't say that. They made that quote up. It's a paraphrase, not a quote, but it's in quote marks. And more confusing, the ending words, "... reduced, the opposite pattern occurred.", are a quote.

It's cool that they are linking a source, but they need to learn to actually take quotes from the source, instead of fabricating them.

Instead of misquoting the study, the WSJ should have tried thinking about the study. For example:
Data from a large real-world social network, collected over a 20-y period suggests that longer-lasting moods (e.g., depression, happiness) can be transferred through networks [Fowler JH, Christakis NA (2008) BMJ 337:a2338], although the results are controversial.
The WSJ could have questioned the wisdom of letting these researchers toy with hundreds of thousands of users in order to produce a paper with a grammar error in the abstract. There should be a comma after "period". This isn't a minor point. The sentence would be confusing enough with the comma, and is harder to understand without it.

On the one hand, I wouldn't expect a publication that misquotes papers (which they could trivially copy/paste from correctly) to notice this. But on the other hand, I don't think they should report on things they don't understand.

Or here is a part of the study that maybe the WSJ would understand:
As such, it was consistent with Facebook’s Data Use Policy, to which all users agree prior to creating an account on Facebook, constituting informed consent for this research.
Instead of misquoting, the WSJ could have accurately quoted this part (it's not very hard, I used copy/paste) and questioned whether it's really "informed consent" when most of Facebook's users have never read Facebook’s Data Use Policy.

How can people give informed consent to something they haven't read? That's the sort of issue newspapers are often better at discussing.

Or maybe the WSJ could put their efforts towards useful commentary on this part, instead of lying about what the study says:
First, because News Feed content is not “directed” toward anyone, contagion could not be just the result of some specific interaction with a happy or sad partner.
The WSJ could have pointed out something interesting and useful here. They missed the opportunity to mention that this is completely false – some News Feed posts are directed at specific individuals. I rarely read Facebook, but I've seen people post stuff directed at a specific individual (this shouldn't be particularly surprising). (How many? I don't know. The study doesn't know either, they just stupidly assumed none are. Apparently Facebook is too far away from their ivory tower to ever read anyone's News Feed.)

There's so much great stuff to discuss here, but the WSJ would rather destroy their own credibility than provide useful commentary.

The WSJ did try to say something worthwhile, but they messed it up. They wrote:
The emotional changes in the research subjects was small. For instance, people who saw fewer positive posts only reduced the number of their own positive posts by a tenth of a percent.
Looking at how big an effect we're talking about is important, and helps put the study findings in context for readers. However, this is factually incorrect and not what the study says. It's bad scholarship again. The study actually said:
When positive posts were reduced in the News Feed, the percentage of positive words in people’s status updates decreased by B = −0.1% compared with control [...]
People who saw fewer positive posts reduced their own positive posting by 0.1% more than the control subjects did.

The WSJ should try hiring people who know how to read and understand studies – and who don't fabricate false quotations – if they want to report on studies.

Note: The article provides a contact email. On 2014-07-02, I explained the two clear factual errors (fabricated quote and misunderstanding of what paper said) and asked about fixing them. No reply. (I'll update my post if I receive a late reply.)

Providing a contact email implies being open to discussion and correction. It implies there is a path forward. If one isn't actually willing to make corrections, it's dishonest.

Elliot Temple on July 5, 2014

Comments

What do you think?

(This is a free speech zone!)