NYT: “Man Bites Dog”; WSJ: “Dog Bites Man”

Remember yesterday’s item about how the N.Y. Times and Wall Street Journal published their very, very different editorials on John Bolton’s resignation? In answer to a Brazilian commenter’s question, the Times is the liberal paper (and therefore anti-Bolton) while the Journal, and most famously its editorial page, is the conservative one.

In today’s papers, there is evidence that those irreconcilable differences also visit the news pages on occasion. Each paper today published an article on a new F.D.A. study measuring the possible increase in suicidal tendencies among young people who take antidepressants. Keep in mind that in the supposed liberal/conservative news split, the conservative Journal is considered to be much more pro-business than the Times.

Here is the Times‘s headline: “Study Finds Medication Raises Suicide Risks in Young Adults.”

And here is the Journal‘s: “Suicidal Thoughts Seem to Abate With Age of Antidepressant Users.”

I don’t want to get all anti-Big Pharma on anybody, but does the Journal‘s headline strike anyone else as a little, um, protective?

These headlines represent a type of editorial interpretation that I am guessing is hard to measure in academic studies of media bias. But to the typical newspaper reader, such differences are so plain as to be nearly comical.


John Fembup

"I don't want to get all anti-Big Pharma on anybody, but does the Journal's headline strike anyone else as a little, um, protective?"

Um, no.

But . . .why, Steven, does it occur to you to question the Journal's state of mind, and not the Times? You wouldn't be a little um, protective of the Times would you?

You are correct of course, these biases are so plain as to be comical.

Anonymous Coward

Well, it depends on what the study actually said. Let's assume that Dubner has read the articles and maybe even looked at the study and his summary, "Each paper today published an article a new F.D.A. study measuring the possible increase in suicidal tendencies among young people who take antidepressants" is accurate. If that is true, then I'd say the times headline is neutral.

donaldo

Come on. There's no getting around that the Journal has used some savvy wordsmithing to actually make it seem as if the drugs are helping when, in fact, the study is stating that they may have a negative effect on young users. Whether or not this was done intentionally we can only debate, but it's very clear that the headline is confusing.

synapticmisfires

Wow, that's really quite remarkable. I suppose the stories could be more similar than the headlines indicate, but those really do tell two almost opposite tales. Incidentally, based on what I've heard, the Times headline is closer to the story, but the contrast alone is remarkable for a regular news story and not an editorial.

jeffstier

The difference in coverage stems from each side's perspective on how to weigh the risks and benefits of pharmaceuticals.

I have sent "Weighing Benefits and Risks in Pharmaceutical Use: A Consumer's Guide," by the American Council on Science and Health's to the New York Times science editors.

A PDF of the peer-reviewed report is available to you, free of charge, here:

http://www.acsh.org/docLib/20050923_bR_web2.pdf

John Fembup

I dount anyone here has read the original study, including Steven. donaldo, you come close to saying that you have read the study, but don't quite come out and say it. But everyone comes to an immediate opinion about which headline is the one that was spun. That's a perfect corollary to the point Steven illustrated, i.e., biases exist in the media, and they exist in their audiences also.

BTW I think it's possible that both headlines can be supported based on findings in the study.

In this case, I also think Steven is being a bit um protective of the Times, because he cashes occasional paychecks from the Times.

Fever

I think this discussion misses the point of media bias in the newsroom. The Wall Street Journal is a business paper and accordingly will print stories that more often are favorable to corporate America. The New York Times is supposed to be an unbiased reporter of news. The content within the articles is the most subtle aspect of their bias, how about their choice of captions or the decision to bury damaging stories in the back of the paper and place strong liberal stories on the front page. For an example of what I'm talking about check out this caption:

http://gatewaypundit.blogspot.com/2006/08/ny-times-hezbollah-photo-dust-up.html

I mean, just wait until the 2008 election and look at the captions selected for the Democratic candidates and compare them to the captions for the Republican candidates.

bk

Does anyone know whether this study has managed to solve the problem of biased self-selection in the sampling? After all, how do you get together a control group of comparably seriously depressed people and follow them while NOT treating them with some sort of SRI?

bk

The ACSH? Let's recall that he who pays the piper calls the tune.

In case anyone is interested in the American Council on Science and Health, don't go look at their web site to find out who funds them. If that information is included on their web site, it's been quite cleverly concealed. To hear them tell it, it's just a group of concerned scientists forming the best of unbiased judgements about science and public health issues.

A tiny bit more digging reveals that it's primarily funded by the industry groups it defends. Doesn't mean they are wrong, but it sure does light my caveat emptor bulb.

c35

I know quite a bit about this. First to get FDA approved, these companies need to run double-blind placebo controlled trials -- so there are no self selection problems. These are well run RCTs.

Second in BOTH arms (placebo & antidepressants) do you know how many suicide in non-adults were observed in the FDA trials with a few thousand patients? ZERO. That's right, zero.

Suicide is a rare event. So instead of using suicide as an outcome, these studies use "Suicidality" or "Suicidal thoughts" as an outcome. There is very little literature on how good a proxy "suicidal thoughts" are for suicide itself.

So the WSJ headline is more correct -- suicidal thoughts is the clinical outcome measured. In theory Suicidal thoughts are correlated to "suicide risks" (from the NYT headline) but there is very little data that actual shows that's true! Therefore the NYT headline is more sensational -- which is what good headline should be. But it should also be true. There isn't enough data to declare the NYT headline is true.

The WSJ headline is also more descriptive in that the differences in suicidal thoughts between the treatment and control groups shrink with age.

So this is not corporate bias. Chances are the headline writer or the author in the WSJ understood the clinical publication it was reporting on better than the NYT headline writer/reporter did.

Read more...

El Christador

Within the text of the NYT story, the story itself indicates that the WSJ headline is true and the NYT headline is false.

In a long-awaited analysis, health officials reported yesterday that antidepressant medications appeared to increase significantly the risk of suicide attempts and related behaviors in adults under 25, while reducing such risks in older people.

The analysis, the most comprehensive and rigorous to date, found that suicidal behavior of any kind was rare, and that people taking the medications were no more likely to kill themselves than those taking placebo pills.

Thus, it is only the risk of suicidal attempts and related behaviors that increase in young adults, not the risk of suicide itself. And the WSJ is absolutely correct.

El Christador

I don't think we can blame the writer of the NYT article for this one. The article makes it clear that the study did not evaluate suicide risk. I think the person who made the mistake is whoever wrote the NYT headline, who did not understand the NYT article.

In general, I find it is fairly common for newspaper headlines to differ appreciably from the actual content of the stories they headline.

SteveSailer

"Man Bites Dog" is an important bit of shorthand for understanding the news side of the newspaper business (i.e., newspapers print stories about unusual happenings, so that the picture you get of how the world normally works from the front page is frequently 180 degrees wrong). Using it to describe the editorial pages diminishes the usefulness of the phrase.

You'd do better to use something like NYT says Tomato, WSJ says Tomahto

John Fembup

Here is the first paragraph of the story from one of the newspapers. Which newspaper?

"The Food and Drug Administration said antidepressants appear to increase the risk of suicidal thinking in young adults but that in older adults the risk declines."

Cam Beck

What struck me about the Times article is the post hoc fallacy of interpreting the data. The Journal (which, contrary to your observation, aside from the editorial, is not conservative) was more cautious in interpreting the data, which was actually the correct way to say it. The conclusion is not definitive, and it relies on someone analyzing the data. There is a correct answer, but it does not logically follow that because two things are correlated, that one causes the other. I find the Times article more misleading than the Journal's.

John Fembup

"I don't want to get all anti-Big Pharma on anybody, but does the Journal's headline strike anyone else as a little, um, protective?"

Um, no.

But . . .why, Steven, does it occur to you to question the Journal's state of mind, and not the Times? You wouldn't be a little um, protective of the Times would you?

You are correct of course, these biases are so plain as to be comical.

Anonymous Coward

Well, it depends on what the study actually said. Let's assume that Dubner has read the articles and maybe even looked at the study and his summary, "Each paper today published an article a new F.D.A. study measuring the possible increase in suicidal tendencies among young people who take antidepressants" is accurate. If that is true, then I'd say the times headline is neutral.

donaldo

Come on. There's no getting around that the Journal has used some savvy wordsmithing to actually make it seem as if the drugs are helping when, in fact, the study is stating that they may have a negative effect on young users. Whether or not this was done intentionally we can only debate, but it's very clear that the headline is confusing.

synapticmisfires

Wow, that's really quite remarkable. I suppose the stories could be more similar than the headlines indicate, but those really do tell two almost opposite tales. Incidentally, based on what I've heard, the Times headline is closer to the story, but the contrast alone is remarkable for a regular news story and not an editorial.

jeffstier

The difference in coverage stems from each side's perspective on how to weigh the risks and benefits of pharmaceuticals.

I have sent "Weighing Benefits and Risks in Pharmaceutical Use: A Consumer's Guide," by the American Council on Science and Health's to the New York Times science editors.

A PDF of the peer-reviewed report is available to you, free of charge, here:

http://www.acsh.org/docLib/20050923_bR_web2.pdf