"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?
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". You're correct that if each answer of a four part multiple choice question is selected by exactly 25% of the test takers, then you have evidence that the test takers knew absolutely nothing and just randomly guessed. But most questions make some sense to most students, and they'll select the most likely answer, and that question will have a higher response rate than the rest. In question K09 at http://christianparty.net/timssk09.htm , answers A, B, and D were selected 11-17% of the time by both sexes, and question C 40% of the time by both sexes, so C was clearly the correct answer. But when only 28.6% of American girls got it correct, and 50.7% of boys got it correct, then you know that few if any girls understood the problem.
They COULD have thought they had the right answer when they answered incorrectly, except that this would require us to believe that close to 25% thought it was A, close to 25% thought it was B, close to 25% thought it was C, and close to 25% thought it was D.
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.
>> >So based on this educated guess, P2 has to be 5.1%.
> >
> >What this means is that no girls knew the correct answer [P1 = 0] and 5.1%
> >of them thought they knew the right answer but had the wrong answer [P2 =
> >5.1%].
> >
> >It's possible but highly unlikely that P1 > 0, in which event we'd have to
> >subtract P1 from P2 to know exactly how many of them were this misinformed.
>> You have no basis for determining either P1 or P2.
>
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%.
> >So the question is, Cary, why were American girls so misinformed about this
> >particular problem (and a whole host of other problems which had a similar
> >result)?
> >
> >The way I see it, there are only three possibilities:
> >
> >1) They learned the wrong thing in the classroom.
> >
> >2) They were taught properly in the classroom, but relied on their
> >"intuition" rather than on what they were taught.
> >
> >3) The error in the test is greater than 3% (like around 5%).
>> There are a multitude of other possibilities.
>
No there aren't, and if you really thought there were, you would have, and *should* have, posted them.
> >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 http://christianparty.net/timssfullreport.pdf 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?
>> >If they were taught the wrong thing in the classroom, then why did boys, who
> >scored 27.2%, get taught the right thing? 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.
> >
> >And that leaves only one possibility: women's intuition. The way the
> >collective feminazi mentality on this forum always manages to produce
> >exactly the *wrong* answer is a great demonstration of what happens when
> >women rely on "women's intuition" rather than facts.
>> 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.
http://christianparty.net/timssphysics.htm
> >American girls got a much lower score than if they'd just guessed,
>> 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?
John Knight
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://iubio.bio.indiana.edu/bionet/mm/neur-sci/attachments/20020718/addbaebc/attachment.html
