I didn't get that from the article and I don't see why you should presume that based on a popular press news article. The article seems to describe a bunch of techniques that they use to ferret out cheaters but each technique (which surely isn't an exhaustive list) surely isn't used on each test.

The metric of large gains is best used on tests like the GRE or LSAT where there is sufficient incentive for the test taker to be halfway competent to begin with so future scores are unlikely to be extraordinarily higher; contrary to what some believe, there are marginal returns on studying for tests like the GRE and LSAT. This is less effective for say a 8th grade remedial math test.

Whereas the metric of erased answers is useless for computer based tests like the GRE and much less useful on tests like the SAT where there isn't sufficient incentive for an individual to go behind the test-takers and change answers. There may be for teachers at elementary, middle and high schools where their merit pay can be linked to some test performance measure.

I agree that calories are calories

"A calorie is a calorie" violates the second law of thermodynamics.

The Mineral Miracle? Or a Massive Information Operation?.

The biggest tragedy (to me) about Africa is that the people there are left hanging between the two systems - hunter gatherer and civilized. They cannot go back, because due to our "humanitarian help" their numbers are too high.

Go back to a hunter-gatherer lifestyle? Do you have any evidence for this? There are very few genuine hunter-gatherer tribes in Africa (Hadza, San, a few pygmy groups, and maybe a handful more). Colonization may've pushed some into an agriculturalist mode of production, but the powerful African kingdoms were doing that well before any European had ambitions of exploiting Africans. And since this story of agriculturalists pushing hunter-gatherers onto marginal land fits with the experience of every other geographic location, I have no reason to think that Africa would be unique and that European colonization is a particularly good explanation for this trend.

BTW, after seeing all the posts online about how many people hate the iPad even before it came out should have predicted it wouldn't sell at all - even though it seems to have sold pretty well so far. I guess their theory didn't predict that their theory failed. Oh wait, now that I've posted this it does predict it.

Did you get a representative sample? Or are you just voicing the opinion of a small group of tech blogs?

Science doesn't "change" its mind. It forms models that allow humans to predict how the universe works.

Humor me, but isn't one of the implicit claims of the scientific worldview that if one wants to be accurate in their assessment of the world, that they ought to believe that the current model is true? Minimally, act on that basis. This seems to be a widely held perspective.


Personally, I am pretty sure that isn't how actual scientists actually work (see this book for an easy reference). So if you're the type of person who likes to follow those you are quite clever and accomplished, then that's an objection. Philosophically, science seems to be best fueled, in the normative sense, by inspiration and intuition, beyond mere minor extension of a current theory. The key element to the enterprise of science is that those intuitions are tested in a rigorous manner to see if they correspond to reality.

I'm pretty sure he's toying with the notion of supervenience (Stanford Encyclopedia of Philosophy entry, Wikipedia). Where, it is thought that you can completely explain changes in the more complicated phenomena (people, organisms, molecules) in terms of changes in the less complicated phenomena (basic particles). It may be needlessly complicated to do so. To express the changes in physical system of an animal running using only particle physics may require more time than is left in the life of the universe. But the thought goes, that it can in principle happen.

Maybe you have a more nuanced objection to that view, but that's different. The view isn't as implausible as you make it seem.

Less crime and less punishment by Mark Kleiman is of interest here.

everything degrades in quality over time, and that's just the way it is, and always will be

Everything good will likely degrade in quality over time. Regression to the mean, combined with the fact that creators of these works have lives. Their attention and dedication can fade because they have other things that they care about (different projects, families, etc). I am upset when an author mails one in, but I try hard to compare the work they produced to the alternatives that I would've considered. Usually, even then, the author that I read for legacy's sake comes out ahead.

There have been reports which say that places like Atlanta are still paying for the olympics.

I was under the impression that Atlanta was one of the success stories of economic growth resulting from hosting an Olympics. This story indicates that there were net economic benefits from hosting the Olympics, but you are generally right, the economic benefits from hosting the Olympics are questionable in general. As others have mentioned, Montreal only recently finished paying off the Olympic stadium they constructed for the 1976 games. The Birds Nest stadium that the Chinese were so proud of is scheduled to host 1 event in 2009, but using Beijing as an example is dubious since it seemed clear from the beginning that the Chinese intended to host a hugely wasteful Olympics for ego purposes.

I am quite confident that both disciplines have gaping holes in their respective theoretical frameworks. I do not see how this entails that an amateur with little to no experience in the field ought to be taken remotely seriously.

Because Slashdot houses people who are clever enough to poke obvious holes in research but either lack the mathematical skills, the energy or the disposition to determine whether their vague objection is meaningful? My personal theory has long been that tech folk are underused cognitively, so they do not gain an adequate "respect" for the difficulties inherent in other fields (this may also due to the type of personality that is attracted to tech-like work). Scan through any Slashdot thread on psychology and/or economics and you'll find droves of people who seem to think that those respective disciplines have overlooked some minor issue. The arrogance is helpful in the sense that you can see through the various falsehoods that are considered true in those disciplines, but it also can contribute to the type of sloppy thinking that you have a problem with.

The next best thing was to give a placebo such that the control group would be confident in their new-found immunity to HIV, at least as much as the experimental group. Otherwise the control would use more condoms because they're not on the experimental vaccine.

This page and this page indicate that the study was double-blind. If it was, then I do not see how your worry is reasonable. If both groups were unaware of whether they received the treatment or not, then I do not see how one group that happened to be the control group would reliably act differently than the experimental group. Am I missing something? Or are you claiming that once people believe they have the vaccine, that they will have more unprotected sex and thus increase their risk of contracting HIV?

Then look at effect sizes if you're worried about gaming statistical significance due to sample size. If you're worried about the file drawer problem, then you're welcome to Google around to see if there are any other trials involving the two drugs used (vCP1521 and AIDSVAX B/E). A search of the US gubment's page on clinical trials reveals none. Also realize, that the big journals require that a study that seeks to be published was registered beforehand, and I cannot imagine that an AIDS vaccine study will be taken seriously if it isn't published in one of the major journals. Though there are issues with this process, it does allow for a systematic review of publication bias rather than a vague reference that is meant to undermine the legitimacy of research.

The advantage that Amazon has over physical book stores is that it can hold practically unlimited number of books. So only now, without the physical constrain, we can practically use top 10% instead.

I am pretty sure a minor but important point made in The Long Tail was that not only did you have to combine massive inventory with a good recommendation system to have a widely successful business. The recommendation system is key since people will tend to regress toward the hits without outside influence.