Junk science from the New York Times

A New York Times article implies that you can improve your health by having more friends:

A 10-year Australian study found that older people with a large circle of friends were 22 percent less likely to die during the study period than those with fewer friends.

However, it seems more plausible to me that healthy people attract more friends than sickly people.

There may also be a social class issue. Upper class people have a larger circle of friends, and it has been well documented that being upper class results in a longer life. This is an example of a confounding factor.

More lousy science reporting from the New York Times.

Comments

[athelas]

I don't see the social class issue necessarily come into play here; it looks like you're just shoehorning your interpretation of this article into your hobbyhorse.

However, you're right, shitty study. People with friends are those who take better care of themselves.

[Peter]

"However, it seems more plausible to me that healthy people attract more friends than sickly people."

It probably works more in the other direction: older people in poor health see their friends drift away because they aren't able to get out much.

[guuy]

What if one has a lot of "bad friends" who are inhabitants of the lower class?

[Kudzu Bob]

And people who are sickly don't get out and socialize much. A slim, well-to-do, gregarious college friend of mine suffered from asthma, a condition that worsened considerably by the time she reached her thirties. The various steroid medications she took for it not only induced hypertension and suppressed her immune system, but also caused her weight to balloon well past the two-hundred-pound mark. By the time of her death from a massive stroke last year at a shockingly early age, she was a recluse.

Does working for the Times turn people into dumbasses, or do dumbasses just end up for working for the Times? Now that's a feature story somebody should write.

[JohnM]

If any you guys bothered to actually read the abstract of the study:"After controlling for a range of demographic, health, and lifestyle variables, greater networks with friends were protective against mortality in the 10 year follow up period." Thus the variable of "health" is already factored out.

The whole "doesn't seem plausible to me = junk science" is the whole reason we have creationists roaming the earth.

If you're gonna trash a study, at least take a minute to read the abstract.

[HS: Sorry, don't trust the researchers to properly control for confounding factors. Seen too much bad research.]

[rightsaidfred]

I wonder if on-line friends will count. Aston Kutcher might live forever.

[Anonymous]

@JohnM. Be sure to read the research approach and data analysis sections as well. Most people screw up either the way they control for variables, or their statistical analysis.

[Kirk]

Let's suppose the article is true: that having friends does lead to a longer life. Just what exactly are we supposed to do about it?

[Mr. Popular]

"Let's suppose the article is true: that having friends does lead to a longer life. Just what exactly are we supposed to do about it?"

Maybe you could start being the kind of person people want to be friends with.

[dearieme]

If it's merely about correlation then they might equally well have said "mortality was protective against greater networks with friends", which must certainly be true in one sense.

[Tanked]

"If it's merely about correlation then they might equally well have said "mortality was protective against greater networks with friends", which must certainly be true in one sense."

If you're the kind of person someone would take a bullet for, I'd say you're set. And how's that unpasteurized cheese treating you?

Also, an unrelated, but actually important scientific development that the NYT failed to cover: http://www.bakonvodka.com/

Oh century, oh sciences what a joy it it to be alive! Yea, verily Brothers we are living in great days! You want to know what that shit it? Fuel for the revolution! Onward to victory!

[Peter]

Assuming the reasearch actually does control for health and other demographic variables, its conclusion (more friends = longer life) could be quite important. It would show that a better mental outlook can increase one's lifespan, or perhaps that a bad mental outlook decreases lifespan. Significant stuff.

[JewishAtheist]

LOL! This post so perfectly sums up HS:

On the one hand, we have "A 10-year Australian study found..." On the other, reverse causation "seems more plausible" to HS. Therefore, HS is obviously right. Scientists are obviously too stupid to have ever heard of confounding factors.

Oops, a commenter points out that the study already corrected for confounding factors. Well, that doesn't matter, because HS "doesn't trust" them. He's "seen too much bad research."

LOL. Talk about truthiness. Nothing is true unless it agrees with HS's gut. HS's gut feeling based on a five minute read of an article about a study is waaaaay smarter than the professional scientists who carefully constructed and analyzed a TEN YEAR study of 1477 people.

Your blog is called "Half Sigma," supposedly alluding to statistics. Perhaps you should rename it something more to do with feelings, like "Half Instinct."

[HS: I subsequently read the research paper online, and it's clear to me that confounding factors are far from properly accounted for. Nor is the raw data publicly available for people to play around with themselves. This is junk research.]

[Underachiever]

HS,
how did they fail to account for the factors?

I have a sneaking suspicion JA is right on this one.

[Brutus]

As a former psych student, I can definitively report that mentally ill people have few, if any, friends.

The obverse is not necessarily true.

[Curtis the Stunt Bum]

"As a former psych student, I can definitively report that mentally ill people have few, if any, friends."

Some of them have quite a few. You and I just can't see them...

[Duke Of Earl.]

I'll withold judgement on weather this is junk science or not, but the NY Times spin -"they overlook a powerful weapon that could help them fight illness and depression, speed recovery, slow aging and prolong life: their friends." is excessive.

1. From the abstract - "Participants: 1477 persons aged 70 years or more living in the community and residential care facilities."

2. Nothing in the abstract indicates how friendship increases survival rates. Note these people are over 70, it may be that having friends around to call 911 every once in a while explains the difference. I don't see anything here that implies friendship fights illness in itself, much less "slow aging."

3. You aren't going to sell newspapers with well hedged but accurate statements though.

[lemmy caution]

Here is the study:

http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1757078&blobtype=pdf

They did do some demographic controls but the cutoffs were pretty low. Do you make more than the standard pension?(they were studying 70 year olds) Did you go to high school? That kind of thing.

I do think that it is more likely than not that a social network helps you live longer. Interestingly, a network of relatives doesn't help you live longer.

[The Undiscovered Jew]

Sigma,

Could you please put up an entry on the recent Miss USA catastrophe?

We need answers such as:

Why was a mentally unhinged homosexual blogger allowed to judge a beauty pageant in the first place?

Does this national disaster have any lessons for HBD/Sociobiology?

I mean, WTF was that all about?!

http://www.youtube.com/watch?v=klT0aSRTuDQ

[100 Miles and Running]

Some homo wrecks the Miss America context? What a shock...Meanwhile:

http://www.butasforme.com/2009/04/20/obama-and-congress-poised-to-end-tax-free-internet-shopping/

We are going to be taxed to death, nickel and dime style. Expect a VAT tax too. You heard it here first. All that bullshit about not taxing the middle class was and is exactly that: bullshit. That's where the money is and that's who Obama and the liberals are going to rob. The only bright spot is that the liberals and hipsters who thought this guy was so great are going to get hit too. Time for the underground economy...

[The Undiscovered Jew]

"Time for the underground economy..."

It's time for the underground, period.

[Turambar]

HS's accusation is that this is more "lousy science reporting" or "junk science" from the NYT.

And indeed that is what Tara Parker-Pope is known for.

This article is a blunderbuss of quotes of academics handwaving, allusions to various studies and citation of factoids intended to pass for a logical argument.

For instance study participants grossly over-estimate the steepness of a hill. When they are paired up their estimate is slightly better.

The conclusion that Tara Parker-Pope leaves the reader with is:
"People with stronger friendship networks feel like there is someone they can turn to,” said Karen A. Roberto, director of the center for gerontology at Virginia Tech. “Friendship is an undervalued resource. The consistent message of these studies is that friends make your life better.”"

Is that justified? Apparently comparing ridiculously poor estimates of hill steepness with a partner improves them, but even if you accepted that having a friend group leads to a longer life does it lead to a better life? Since the survivors in the Australian study were much more likely to be married and in a "residence community" maybe the study principally means that the elderly can extend their lives by having someone around to rescue them if they have a slip and fall or a seizure of some sort. That says nothing about the quality of live the average reader.

Even if you like the Australian study you have to admit the Times article way oversteps- which is what HS was critiquing.

[Jeremy]

"A 10-year Australian study found that older people with a large circle of friends were 22 percent less likely to die during the study period than those with fewer friends."

The above excerpt in no way implies that those living longer are doing so BECAUSE of the larger social circles. Just that they were less likely to die in that period. There is no causal relationship implied, at least by the above statement, they are merely saying a correlation exists.

[Duke Of Earl.]

That statement didn't imply it, the NY Times article did, it sells better that way.

[Andreia Stelet]

I think the study refers to real friendship. People you can rely on and share your life. I doubt anyone can have more than 5 in a whole life. A have a few, very few. I treat them like gold. And they also know they can count on me at anytime for anything. Unfortunately there are many people nowadays that don't know the difference between friendship and social networking. My social network has around a hundred of people. I also value it of course.