brain sizes: Einstein's and women's
Bob LeChevalier
lojbab at lojban.org
Fri Jul 19 03:08:41 EST 2002
"John Knight" <johnknight at usa.com> wrote:
>"Bob LeChevalier" <lojbab at lojban.org> wrote in message news:sgnbjukhqp962shh6oiqtb12ej3naeh321 at 4ax.com...
>> "John Knight" <johnknight at usa.com> wrote:
>> >It's impossible to know *exactly* how many girls thought they knew the
>> >answer,
>> >and instead had the wrong answer (P2). But because so few got the correct
>> >answer, it's a good educated guess that P1 can't be greater than 0.
>> >
>> >In the example above where 14.9% of girls got the correct answer, we know
>> >that 85.1% of them got the wrong answer, and that they did so by guessing
>> >wrong.
>>
>> No we don't know that they "guessed" wrong. We merely know that they chose
>> the wrong answer. There are a multitude of wrong answers that can be
>> obtained by faulty logic, erroneous assumptions, misreading the question, or
>> even erroneous teaching. TIMSS people chose specific wrong answers to catch
>> the most likely of those errors. "Guessing" has to be RANDOM in order to use
>> probability to predict the distribution.
>
>Sheesh, I can't even imagine whoever taught you all this garbage. The
> schools in jew Yawk are even worse than their miserable test scores indicate
> they are, aren't they?
How would I know? I've never attended a school there.
>If a test taker doesn't know the answer, but has read the problem and
> selected the answer which he "thinks" is the best one, then the pattern will
> be everything but "RANDOM".
If that is the case then all of your statistical gibberish about 25% is
hogwash, because simple probability analysis only applies to completely
random processes.
>How could each wrong answer have been taught as the correct answer to 25%
> segments of the girls who took the test? It's impossible, so we know that a
> large number of them guessed randomly, rather than thought they had the
> right answer and didn't.
Thereby revealing another misconception about the test and about American
education. American teachers did not teach correct answers to the test; nor
did they teach incorrect answers to the test. American teachers (nor any
other teachers) had no idea what the test would consist of, and they taught
NO answers to the questions on the test. Indeed, one reason why Americans
scored low on TIMSS is BECAUSE so much of the material on the test is not
taught at all in our schools, since it is not part of the mandated curriculum
of any state.
>You're dead wrong. When so few test takers get an answer correct, but 0% of
> the questions were "not reached", then we know that P1 on question K09
> cannot be higher than 3.6%.
False.
>> >Based on the repeatability of the test, and its relative consistency with
>> >other tests given by different organizations, I believe that their estimate
>> >of a 3% error is very conservative [read: too high].
>>
>> There is no 3% error. The TIMSS numbers are exact, and not statistical
>> samples. You pulled the 3% out of your strange orifice.
>
>WHAT? You really know nothing about this, do you? Why do you bother to pretend that you know something? Read
[cite deleted]
> in its entirety if you want to understand how wrong you are. Pay particular attention to Table A2.2 which shows the standard errors for advanced math scores. And I quote:
>--------------------------------------------------------------------------------
>
>The 95 percent "plus or minus" confidence interval around each nation's score is two times the standard error.
>
>TABLE A2.1
>
>NATIONAL AVERAGE SCORES AND STANDARD ERRORS:
>
>MATHEMATICS AND SCIENCE GENERAL KNOWLEDGE
>--------------------------------------------------------------------------------
>You will note on that table that the standard error for the US, who scored dead last at 442, was 5.9. Multiply 5.9 by two, and divide by 442, and you have 2.7%. How can you possibly not know this?
I know this, but you seem to have confused "standard error" which was 5.9,
with the 95 percent confidence interval, which was double that amount.
But whichever you think you were referring to, that is the error on the
NORMED SCORE of 442, which itself is a statistical calculation. It is NOT
the error on the source data for the individual test questions. If they say
28.6% got the answer correct, that is the actual percentage and not subject
to the same error analysis as the overall normed score.
>> >If they were taught the wrong thing in the classroom, then why did boys, who
>> >scored 27.2%, get taught the right thing?
They didn't, if only 27.2% got it correct.
>> >Of course it's possible that they
>> >too were taught the wrong thing, then realized during the test that what
>> >they were taught was in error, and adjusted their answer accordingly. Of
>> >course, if this is true, still far too many of American boys got it wrong to
>> >believe they were even taught this at all, particularly since 60% of Swedish
>> >boys got it right.
>> Women's intuition does not explain D12, where men scored less than 17%
>> correct. Furthermore, if women's intuition were just another form of guess
>> then it would be random.
>
>You obviously didn't understand the point about women's intuition. The only
> way women's intuition could influence scores so negatively is if the girls
> remembered enough about the subject to sense what the correct answer is, but
> then answered it wrong, subconsciously, or almost intentionally. The simple
> fact that girls scored consistently lower than they would have if they had
> just guessed on ONE THIRD of the TIMSS physics questions for which we know
> both the problems and the scores is PROOF that "women's intuition" is a
> fraud, and that the joke is on women.
Your analysis does not carry over to women of all countries. Nor does it
explain why nearly as boys had nearly as many problems where they scored
lower than if they had guessed.
Your explanation is simply out of touch with reality.
>> On a couple of problems, fewer girls chose the correct answer than if they
>> had guessed randomly. Likewise for boys. It is likely that someone guessing
>> on the entire test would have scored 200, which is typically the lowest score
>> possible on a normed test.
>
>Deal with the percentages, lojbab, that way you won't confuse yourself even
> more than you're already confused. It wasn't just a "couple of problems"
> where girls would have gotten higher scores if they had just GUESSED--it was
> ONE THIRD of one part of this test.
>
>Why is this so difficult for you to confront? What is it about the TRUTH that you just cannot bear to FACE?
The truth is that you are a nincompoop who does not know what he is talking
about. I have no trouble confronting that truth, and for that matter, in
confronting you whenever you abuse that truth.
lojbab
More information about the Neur-sci
mailing list