[lbo-talk] Maximise or satisfice?

ravi gadfly at exitleft.org
Wed Sep 29 09:40:18 PDT 2004


Miles Jackson wrote:
> On Wed, 29 Sep 2004, ravi wrote:
>
>>of course a lot more analysis is needed to show that IQ test results and
>>academic performance, even when they correlate, are not tracking
>>variables (perhaps of a trait like aggression). there is also the
>>trivial danger that one or the other result (more probably performance
>>in intellectually challenging areas) is manipulated or ill-deserved.
>
> Let me get this straight: aggressive people somehow get higher IQ
> scores and better school outcomes? whaaaa?
>

the bit about aggression is in paranthesis and prefaced by "perhaps". nonetheless, its possible that aggressive people do better at competitive tests, into which category both entrance tests and academic tests fall. that is not sufficient reasoning to establish that point, but my real point is that IQ test performance and academic performance may correlate well because they do so with another root-cause variable. etc. etc. as you probably already know, this is all fairly basic statistical analysis issues.


> It's not that complicated:
> IQ tests predict objective, nontrivial outcomes like completing
> medical school, getting a college degree, passing the bar.

i guess i will have to hit those psych 101 books you recommend. a quick bit of googling produced no substantial data. what you suggest above doesn't sound unreasonable to me.


> Are you
> really so cynical about formal education that people don't actually
> learn anything in any of this professional training? (And if you
> think these are ill-deserved credentials, why do you rely on the
> services of computer programmers, health professionals, and
> lawyers?)

its a distraction to get personal, no? i am a computer programmer myself, and a product of formal education. its doubtful that any of the stuff i learnt in school or college is particularly of use to me, right now, but all that is anecdotal and useless.

i am not cynical about formal education. people, i am sure, learn various useful things in their training. that's an absolute issue of the value of the education system, isnt it? what i am in search of is the particular information on relative performance and its correlation with IQ or entrance testing (i admit this particular thread is not about entrance tests).


> It's strange that in the context of this list, I appear to be
> a supporter of IQ testing.

why do you find it strange? it seems in keeping with leftist positions to thoroughly question and examine any objective unitary ranking system that could (and probably does) affect the lives and opportunities of individuals.


> In fact, I'm a pretty vociferous
> critic of the misinterpretation and misuse of the tests.

which is probably the same thing that motivates the responses to your posts. note that at least i (among the responders) do not disagree with your point. wojtek [sp?] for example, posted recently on the dangers of false negatives. also, see below:


> But
> c'mon, let's base our arguments on evidence!

which is exactly my point. hence my request for pointers to data. as you suggest, i could go read psych 101 texts, i suppose.


> (I should note
> that the problems that ravi points out about the SAT do not
> exist with IQ tests like the Stanford-Binet.)

as pointed out by me (including in the quote from MSN encarta).

--ravi



More information about the lbo-talk mailing list