On Fri, 11 Jun 2004, snit snat wrote:
> I think the problem is this: he's saying it's not an epidemic. You know
> what an epidemic is--everyone is catching it. He's saying that the
> numbers suggest that the fatsos are just getting fatsoier, not that
> everyone is getting more fatso (or catching it).
Right. But he's wrong. If he was right, the average would change but the median would stay the same. Instead, the median has changed, as he himself admits.
The way in which he dismisses the change in the median shows he has no knowledge of what median means or of what counts as statistical significance. He must, you'd think, in his own microscopic work. But this is what comes of having no practice in macro statistics. You can make the most elementary mistakes and not know it.
That dismissal is crucial to his argument. If it's wrong -- and it is -- nothing else can follow. He's got nothing to stand on.
A more sophisticated version of his argument would be that the skew has increased -- meaning that whatever is affecting everyone is affecting superfatties more. That could be. You could run those stats yourself if you want. And you'd be more sophisticated than he is. And be contributing something, which he's not.
> He then points out that eating more [calories] than you should by just
> 10% can add up to significant weight gain.
And the statistics show that that is exactly what has happened: the average per capita daily caloric increase has increased 10% over the last 20 years (during which time, btw, Friedman has done exactly zero original research on this subject -- he's been in the lab all that time).
Now for his argument to hold, the 25% who are obese would have to have increased their average intake by 40% and everyone else stayed the same. You think that's true? Go find the stats to support it. He certainly hasn't. He's just riffing off the top of his head.
He's an oversimplifier, kelley. There's no escaping it.
No one will ever accuse you of that, though :o)
Michael