Monday, April 26, 2010

Meta-science Dismissed

I was reading about some reasons to distrust meta-analysis, and when I went to evaluate the ideas' merits, I remembered a source of error that I'd read about recently. I enjoy Eades' epistemological takedowns, but this one basically amounts to, "Meta-analyses are usually done wrong."

Eades is correct to dismiss meta-analyses, though perhaps not for the reasons he thinks. (I suspect experience guided him.) Although he has good reasons; a meta-study including methodologically poor studies is going to be a poor meta-study, and a meta-study has to be very careful to include data fairly. If it fails to do either, I would call it a mistake, not a study; it's quite possible for a meta-analysis to do both properly.

This mainly impacts journalistic writing, as they'll happily write on any study that fits their agenda, and generally forget caveats and other subtleties. If you're in a position to read the actual paper, you're in a position to tell if they're cherry-picking data.

However, how many meta-studies do you suppose correct for the fact that neutral or negative studies don't result in papers? I'm not sure how I would check, but I'm going with negative zero. They can't even try to deal with cherry-picking that occurred before the paper is published.

It's a shame. Yesterday, the meta-study was my favourite kind of statistical study.

Thursday, April 22, 2010

Flynn Effect vs. Retards?

The clinical cutoff for the mental cripple is an IQ of 70. Flynn noticed that the average - normed to IQ 100 - is getting smarter. There's a qualitative cutoff at the mental cripple stage. Such a person is literally too stupid to support themselves, and require lifelong care. Basically, childcare extends across adulthood.

Setting aside edge cases (which can be statistically smoothed out) is the cutoff actually dropping, as the Flynn effect would predict?

Saturday, April 10, 2010

An Obvious Dilemma?

Blogging and print journalism seem to share a dilemma.

Obvious thought is not necessary; the audience can figure it out on their own. The only real barrier is inclination, which can be overcome simply by vaguely gesturing in a particular direction.
Non-obvious thought is automatically more useful, but deters audiences, because it is difficult both to convey and to understand.

I'm not interested in reading or writing about the obvious, and you're reading this blog; you're probably the same way. There don't seem to be enough of us to fully support blogs that pursue non-obvious ideas.

This can somewhat be overturned by posting long enough to put forth a full explanation. Beyond generating complaints, I wonder what effects long posts actually have on readers. Books work; is there some no-mans-land in the middle?

I wonder if a compromise would work. Perhaps consistent posts on the obvious peppered with eyeball busting treatises, at say 10-1. I suspect there's a lot of intersection between readers who want to be in the choir and readers who want an uphill read.

I think this dilemma has defined the idea of 'news.' One way to be both simple and useful is to supply small updates to existing bodies of knowledge. Unfortunately for the purposes of discourse, small and obvious amounts to insignificant. Thus, neither actionable ideas nor mind-changing ideas are economically viable to share. My issues with journalism may stem from journalism bloating beyond areas like weather, where facts that are insignificant as ideas can have significant physical consequences.

The part about 'existing knowledge' seems like it would drive expectations in the wrong way, too. The existing knowledge is necessary to make the update obvious, but having been raised on a diet of the obvious, can you deal properly with the non-obvious? Isn't your first instinct going to be to treat it as obvious but wrong? It was certainly my first instinct, and it would rather neatly explain around 3/4 of blog comments.

This dilemma has probably shaped the idea of 'science' as well. Specialists cannot be economically sustained by selling the ideas they develop; hence grants. Given the abysmal efficiency we're seeing grants achieve, I hope this one is a false dilemma. Since selling to a journal-equivalent is not sufficient, science should find some other voluntary exchange which does supply enough wealth. (Indeed, science did not begin in the age of grants, though measuring science by number of papers began then.)

On the other hand, sharing ideas is something humans seem to do for free. From time to time. Experiments remain expensive, but I have to ask; is there demand for science? If grants weren't cornering the market, would there be efforts to find voluntary alternatives, or would science die out, as grant-proponents (naturally) say?

More importantly, is there actually a market for ideas per se? Funding thinkers certainly puts more ideas out there, and conversely, perverted and neglected IP rights suppress the idea market. But exceeding the actual demand inherently implies waste. If ideas cannot be sold above the cost of production, there are more ideas than people want. Perhaps neglecting the market is actually the good choice. Perhaps the only reason there are so many ideas out there is that one day long ago, a scholar was frustrated that their preferred leisure was unprofitable, and then, through the devil's own luck, came into power.

Tuesday, April 6, 2010

Taxation versus Crime

I was surprised to discover how badly neglected the direct link between taxation and crime is.

Criminals commit crime because they perceive that crime pays. Conversely, the more rewarding legitimate work is, the more limited the rewards of sacrificing legitimacy are, in both quality and quantity. Anything that reduces that reward is going to directly stimulate crime. All this can be derived more or less a priori.

I've only found one place discussing this link, and the data look bad to me. Beyond the simple confirm/deny dichotomy, different types of crime will respond to variance in payoff differently.

This means a few things. Even setting aside anarcho-capitalist flavoured violations of property rights, raising taxes has an inescapable odour of immorality. Even setting aside corruption, government waste not only wastes the honest dollar, it creates new dishonest dollars. (Though I stress that rewards, both honest and dishonest, need not be financial.)

Finally, a patchwork would not end up taxing at the Laffer maximum, because a patch with lower taxation will inevitably have a lower crime rate, thus attracting immigration, all else equal. Competition implies market implies consumer goodies.