Why it's hard to teach evolution

It's funny, I put ten Elvis Costello songs and ten Meat Loaf songs on my playlist and played it all the way through, and it seemd to spend 90% of the time playing Meat Loaf.

And the poor bastards on the iTunes and iPod teams were forced repeatedly to test and retest their random number generator.

Which then leads them to the not-at-all-random Smart Shuffle feature just to shut everybody up.

The moral of the story might be that people don't actually want truly random shuffles. Their expectations are that shuffling will get them even distributions of artists through the mix.

Is it possible the uneducated actually have a MUCH better feel for "random" then the educated? Academia tends to limit acceptable perceptions rather than to expand them. The moment that Gaussian or Poissant distributions parameterise your thinking, the true essence of random is lost.

I agree that Ceej's use of “uneducated” (as if education were a monolith) is a bit snobby and naive—you have to reinterpret it as applying only to a limited type of education and assume that the education took.

On the other hand, every sentence of your post is condescending Po-Mo gibberish. People with some experience thinking about randomicity have a MUCH better feel for it than the uneducated.

Academia tends to limit your perceptions of a thing, perhaps, but it expands your perception of things in general (of course if you have a vocational "education" as from an Engineering school no such assumption).

It's the appreciation of distributions that allows you to understand at all how large repititions of the random aggregate.

Though it's completely true I've seen some "educated" people stake too much on getting an "average" result, I can't say that parameterizing thought or "true essences" are likely to have any actual meaning.

From the biologist's perspective, the idea of randomness very directly interferes with the teaching of evolution, since NO MATTER how many times you explain to a given student, grad student or other biologist that original genetic changes are an entirely random event (selection then being the non-random component)... the very next time you hear that person present on the development of some animal or microscopic organism, they will inevitably get cause and effect completely screwed up or indicate some higher level of organization or intelligence. (the latter being an obvious indication of their own lack.)

Talk to them about it afterward, and half will say "of course, I meant it in the correct way, it just came out wrong" and the other half will argue with you that their explanation was completely in line with the current understanding of The Theory. Which only goes to prove that you can get very very far in the educational system without understanding a single iota of what your memorized 'facts' are based on.

This is maddening. That's why I had to share. Makes me want to inflict violence upon both parties. If you know how it works, say and explain it in truth, especially in front of cell biology students. If you don't know how it works, and beleive darwinian evolution to be the same as the **** of Wallace, go home and stay away from students forever. (and from reproduction in general)

With Great Affection, and a Very Strong Wish that you will one day reproduce in a plentiful way, hollyjean

and what does Po-Mo mean, Zink?

While I wait, I'll just hang out over in the creationist camp and try not to have any of my degrees stripped from me.

(evil laughter bounding off the U of Wisc walls) 'hjj

Generally I mean in the style of those authors examined by Bricmont & Sokal in Fashionable Nonsense: Postmodern Intellectuals' Abuse of Science, but my touchstone is Roland Barthes's Writing Degree Zero.

In general I apply it to people who reproduce the form of intellectual or academic discourse:

• use of logical constructs like "if this then that" or "it then follows"
• use of "academic" words like "parameterise"
• confident statements of assumption or conclusion "Academia tends to limit..."
• novel paradigm shifting conceptualizations "Gaussian or Pissant distributions parameterise your thinking"
• radical ideas—"the uneducated actually have a MUCH better feel for "random" then the educated"

without using it to describe intelligent thought. They are like Macaws mimicking human speech; it's as if they understood what they were hearing well enough to reproduce various verbal structures, but never saw the underlying interelation of elements that gives an argument logical force.

So spoondude reacts to Ceej's jab at uneducated folk with a radical idea: "What if the world were upside down? What if the uneducated know more than the educated?" which carries within it a ton of cultural assumptions (as did Ceej's) and is obviously emotionally motivated, probably some kind of denial.

Since Radical ideas are necessarily true—the idea that one class has non-economic advantages over another is a reflection of the advantaged class's control of discourse and never a reflection of reality—his task is next to uncover why it is true. He's directly ratiocinating. His motivation to defend this class is never at issue, of course. And he's not going to take the discussion too seriously, since he know's it's unlikely any of us can be dislodged from our self-gratifying sense of class superiority.

This sort of ratiocination is incredibly common. If you present a Creationist argument to Ceej she's not going to consider it carefully and thoughtfully, she's going to assume it's BS and probably just ignore you. And it's the same technique Colbert uses to disembowel liberals: he knows they're talking BS, it's just a matter of catching them at something.

Then spoondude launches into full Po-Mo mode. He writes "Academia tends to limit acceptable perceptions rather than to expand them." Now, I can make some sort of meaning out of that, in fact I can make several. I can't make many in which the sentence is true or relevant, but I can see how someone might say them.

In the Po-Mo world such ambiguity is a jewel, a sort of quantum-mechanical super-position of possible meaning. Interacting with a few more heisenberg sentences you get a paragraph which cannot be criticised on its own terms because it doesn't commit to any terms. Any attempt to understand the paragraph without committing to agreeing with it is met by some kind of attack, based on your misunderstanding of some unstated idea.

This is also the basic form of forum conversation since the inception of the internet.

Spoondude's two obvious followups to a meta-analysis like this are:

• actually I was just pulling your leg, it was parody (a-hole)
• you don't understand modern thought well enough to follow my argument (po-mo)

If you ask most people to pick a random number from 1 to 5 they will pick 3 If you ask again, almost nobody will pick the same number twice That is because there is an expectation that random means no pattern

Most people also don't understand odds, combining that they will figure out that 1,2,3,4,5,6 has a lower chance of coming up in the lottery than their more random numbers

Most people have average or lower intelligence

There is no word for unpredictable but evenly spread out selection, so people overuse random. The problem was, the engineers heard random and thought, finally a spec we understand.

Last thought: How do you prove that something is random, without making the same assumptions laughed at here, only with more data points

One generally supports claiming something mimics random by demonstrating that its spectrum mimics that of white noise. Or some other noise depending on the type of random you are doing.

Obviously for a random ordering of N elements, when you have seen N-1, the last is completely determinate.

There are well established methods of evaluating randomness. See DIEHARD (http://www.csis.hku.hk/~diehard/) and or NIST STS (http://csrc.nist.gov/rng/).

It takes great conviction to prove that methods exist of evaluating randomness and cite a link (http://csrc.nist.gov/rng/) that states: However, no set of statistical tests can absolutely certify a generator as appropriate for usage in a particular application

I had pretty much the same thought: Why does it matter if the pseudo-random number generator is truly pseudo-random? It's how you use the random numbers to choose songs and specifically which songs you end up with that should be subject to statistical analysis.

In related WTF, I was at work the other day and they were discussing hashing IDs and assigning the objects to buckets. I said, “well, some of the buckets will be empty and some of them will have quite a few entries...” and the response was that a better hash function would fix that!

I completely agree with you that actually random number generators should be freely interchangeable. The page that claims the application plays a role could be discussing performance/space tradeoffs or whatever: at best it leaves a great deal unsaid.

Most people are not of average or lower intelligence. Almost everyone is stupid almost all of the time. To quote Alfred North Whitehead:

It is a profoundly erroneous truism, repeated by all copybooks and by eminent people when they are making speeches, that we should cultivate the habit of thinking of what we are doing. The precise opposite is the case. Civilisation advances by extending the number of important operations which we can perform without thinking about them. Operations of thought are like cavalry charges in a battle - they are strictly limited in number, they require fresh horses, and must only be made at decisive moments. (Intro. Math. p. 61)

Oh, and back "on thread":

http://en.wikiquote.org/wiki/John_von_Neumann

Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin. -- John von Neumann, 1951