They try to correct for those factors but they can never be sure.

Maybe the biggest epidemiological study is the Nurses' Health Study, which has several thousand participants and has been going on for more than one generation. They've been recording a huge number of personal activities and medical developments. Then they run it through computers to find associations. Then they try to correct for all the factors. Then they do a randomized controlled study to find out if the association was spurious or if it really was causation. They get it right about half the time, which is worthwhile. But you just can't eliminate every possible confounding factor.

True, it's appropriate to start with a correlational study before you go on to a randomized, controlled trial.

This would have been a good study -- if they didn't come to an unjustified conclusion.

We don't know that from this study, because they couldn't control for all the other factors that might have influenced weight.

If they really wanted to find out whether sunlight affected weight, they would have done a randomized, controlled trial.

They would have randomly assigned half the people to getting exposed to sunlight early, and the other half to getting exposed to sunlight late.

Instead, they let the subjects go their merry way and simply measured their exposure to sunlight during the day.

These kind of studies give spurious results. For example, suppose the ones who are exposed to sunlight in the morning are getting up early to start their day jogging.

No, I'm not saying that I try to make each side look equally valid.

I'm saying that I try to let each side make their best case, and let my readers decide.

I'm writing for people who are intelligent enough to know how to evaluate both sides of an argument and come to their own conclusions.

Sometimes it's obvious that one side is lying. Sometimes it's too close to call.

For example, when I was writing about needle exchange programs for IV drug users, I had a stack of well-designed studies published in major medical journals saying that needle exchange programs saved lives, and I could call up experts who would make very persuasive arguments for them.

Then I'd call up some right-wing politician's office and say, "What's your evidence? How do you respond to this article in the Journal of the American Public Health Association?" Let them talk. Sometimes they just make fools of themselves. (Governor Pataki said, "We don't have enough evidence yet. We're studying it." Studying it forever.) Sometimes they really did seem to be well-intentioned people who believed things that were wrong. Sometimes they had actually changed their position.

Of course it's always possible that I could find out that I was wrong.

If you want to know more about the process, look up John Stuart Mill's On Liberty on the Internet.

(Oh yeah, the other rule is, "Always ask, 'What's your evidence?'")

I write about medicine. I read the journals and go to the conferences.

I was passing by New York City Hall (during the Giuliani Administration) and I saw a demonstration by AIDS activists, something that I had been covering. I always like to talk to the real people involved, so I tried to get over to the demonstration.

Giuliani put a locked gate around City Hall. I had to stop by a guard post. I told the guard what I was doing, and he told me I needed press identification. I told him that I should be able to go to the demonstration simply as a member of the general public. But he was an asshole on a power trip and insisted that I needed a press ID. Finally I saw somebody else walk through without press ID, so I just walked through myself.

I later called up City Hall to complain about the guard, and went through a long series of written complaints to supervisors who were perpetually on vacation or had been moved to a different job. Finally the City Hall guards let some politician's friend with a gun into City Hall without screening, and he shot and killed a City Council member. It was no longer a good time to press on with a complaint like that.

I also called the City Hall press office and asked them what the requirements were for a press card. They were actually reasonable as written. The original purpose of a press card is to let you cross police lines during a fire or other emergency, or big events or demonstrations, and they gave press cards to reporters who regularly covered them for news media. Counter-cultural publications like the Village Voice and WBAI-FM got press cards. Less formally, they let the cops know when the reporters were watching so they didn't beat up demonstrators with cameras around. With time, press passes turned into a prestige item that publishers and other freeloaders used to try to get out of speeding tickets, get free admission to the circus, cage free meals at restaurants, etc. You had to fill out a form and apply, documenting that you actually do cover events where a press card is useful. I thought that it might actually make a good story, for the National Writers Union newsletter or someplace, "How to get a police press card."

I decided that I don't need your fucking press card. I can find out enough just by exercising the rights I have as an ordinary citizen, and exercising my willingness to go to jail if that's what it takes, to get my readers the information that they want and have a right to know.

One of the things that always amused me was the outrage of the press (like the New York Times) when the cops beat up their reporters during a demonstration (at the Chicago 1968 Democratic Convention, for example). Why weren't you doing your job of reporting the truth when we were getting beaten up by the cops, in front of your own eyes?

So blogger, shmogger. You don't need a press pass to write journalism. All you need are your rights under the Constitution and the willingness to get beaten up and go to jail.

As someone who made a modest living for 30 years as a "journalist" (or whatever you want to call me), I can summarize the most important thing I learned in 30 seconds:

Every time you attack someone, always call him to get his side.

(Variation 1: Every time you write something that you strongly believe, always call somebody on the other side to find out why they disagree with you.)

That's it. If you follow that rule, you'll always get a decent story.

Me too.

I suspect a lot of people will dismiss it entirely because it was done on mice, and humans are not big mice.

And mouse evolution, since this study was done in mice.

When they replicate this study in humans I'll pay attention.

Humans are not big mice.

We don't have a moral obligation to get into a situation that most of us don't understand, and make things worse by escalating the conflict.

http://www.nytimes.com/2013/08...
Germany Backtracks on Tuition
By CHRISTOPHER F. SCHUETZE
Published: August 25, 2013

(German colleges are now free again, like the Scandinavian countries. Under the German constitution, the 16 state governments control finance and education. A 2005 federal court decision allowed them to charge tuition. 8 states, in former West Germany, did, but it was unpopular and they reversed their policy. Lower Saxony charged €1,000 ($1,300)/year. An economist estimated that tuition caused 20,000 potential students (6.8% of all students) to forgo enrollment in 2007. Denmark, Norway and Sweden have free tuition, although Germany, with 2.5 million students, is the largest. Britain raised its tuition caps to £9,000 ($14,000). In France, most public universities charge a few hundred euros per year, though the grandes écoles are more expensive.)

That's true. I read a classic analysis of inequality in the U.S. (sorry I can't remember the citation), which concluded that for people from the lower classes, a college degree with any major was a guarantee of a professional job. This was based on data of people who were working when the study was done, which was probably in the 1960s. So it was true in the 1950s.

Another problem is that correlation is not causation. The one factor that most strongly correlates with your income is your father's income. In general, rich kids go to college. Rich kids don't become rich because they went to college. They become rich because their fathers were rich.

So George W. Bush was right to spend his time at Yale and Harvard partying, getting drunk, and smoking pot, rather than studying business management and boring old wars.

The vast majority of public schools are run by school boards in which parents can have greater or lesser input.

Some districts have smart aggressive school boards that set standards and make sure their kids get a good education; in other school districts it's all about whose brother-in-law gets the lunchroom contract.

It's small-town democracy. If you don't have good schools for your kids, blame your self and your neighbors. It's your responsibility.

I remember one tech college, RCA Institute, that was actually pretty good. I met their graduates doing pretty good work on pretty good jobs everywhere. Bell Labs used to hire techs from RCA Institute. I met an electrical engineer from India who went to RCA Institute for 6 months to finish off his education. Then he went to work designing IC circuits for the blind. He showed me the first 8086 chip I ever saw in my life.

Then they went into decline. They were going to close down, the teachers tried to make a go of continuing on their own, and it just wasn't working out as a viable model. They changed their name several times. Last time I heard it was called TCI, Technical Careers Institute or something. They had rented a space on 8th Ave. off W. 56th St. in Manhattan, next to McDonald's.

Anybody know more about them? Are they still any good?