"Bob LeChevalier" <lojbab at lojban.org> wrote in message
news:f3bcju8u4kmc3k3engod0496aqc835q0ku at 4ax.com...
> "John Knight" <johnknight at usa.com> wrote:
> >> > The ONLY time you could apply that argument is when a large
percentage of
> >> > them answered correctly, but even then, if 0% failed to respond at
all, then
> >> > some of them HAD to guess.
> >>
> >> Well, that's a reasonable assumption, but so what? Some people probably
> >> guess on every multiple choice question on ever test.
> >
> >You and I agree. LeChevalier is an idiot. He said, and I quote: "we
don't
> >know that people guess randomly when faced with a test question they do
not
> >understand. Indeed, we know that they do not".
> >
> >This is not the stupidest thing he's ever said. And he'll deny even said
it
> >as time goes by. But only a real moron "thinks" like this.
>> I see no reason to deny I said it. I believe it is true. People are
> incapable of behaving truly randomly. More likely, if someone were to
have
> no idea what the answers were, but wanted to guess, they would mark the
> answers NON-randomly, like marking "A" for all of them or repeating a
> pattern.
>
I don't blame you for trying to retreat from your stupid remark, so let me
repeat it for you:
"This makes the assumption that those who know nothing guess randomly. IN
reality, we don't know that people guess randomly when faced with a test
question they do not understand. Indeed, we know that they do not."
Note that you said "we KNOW that they do not" guess randomly.
"WE" "know" no such thing. "WE" know that when students are confronted with
a question which they don't have even a clue about, and they STILL answer
that question, that the only thing they could possibly do is "guess
randomly", but I agree that *YOU* will probably never "know" that.
> >You don't seem to understand the point either. If *all* students just
> >*guessed* on a four part multiple choice question, and didn't have a clue
> >about what the answer was, they would have gotten 25% of them correct
just
> >by chance.
>> But the converse is NOT true, that "if just 25% of them got the question
> correct, then all students just guessed".
>
What kind of a freak are you? It IS true. It's a basic FACT of probability
and
statistics that if you have a four choice answer, and exactly 25% of
students select the correct answer, that this could be nothing BUT a random
selection.
This is particularly true when all four of the answers are as close to 25%
as they were.
If 26% selected the correct answer, then 74% got the WRONG answer, and the
obvious scenario when there is such a low correct response rate is that
students GUESSED. This is too close to the standard error to be
significant, though.
If 30% got it right, then 70% got it wrong, and it's very rare that 70% of
students would have been taught the wrong thing. Certainly it happens, but
if the answers are spread evenly across the other options, then it's
inevitable that many of them just RANDOMLY guessed.
> >The only way for them to get less than 25% correct would be to know
> >something about the problem and select the wrong answer on purpose, or to
> >have the wrong information in the first place.
>> A slight dawn of understanding. Of course when they "select the wrong
answer
> on purpose", they don't think it is the wrong answer.
>
No, there was no dawn here. There would have been a dawn if you'd realized
that American girls literally cannot score lower than if they'd just guessed
on a THIRD of the problems unless they knew enough about the subject to
subconsciously select the wrong answer (unless they were taught wrong in the
first place, which is silly, because they were taught in the same classroom
as boys who didn't do this). http://christianparty.net/timssphysics.htm
On one or two questions we could chalk it up to an unknown anomaly. On a
THIRD of the questions, you have a condition that can't just be swept under
the carpet.
> >If 30% of them got it correct, this doesn't mean that 30% of them knew
the
> >answer.
>> It might or it might not.
>> If they didn't have the wrong information, or didn't make an error,
> >then of the 30% who got it correct, 23% would have gotten it correct
because
> >they guessed, and only 7% would have gotten it correct because they
> >understood the problem [ x = total guesses, 0.25x = correct guesses 0.75x
=
> >wrong guesses = 70%, x = .93, 0.25x = .23 = correct guesses, correct
total
> >answers of 30% - 23% correct guesses = 7% (those who knew the problem)].
>> You cannot determine the percentage who guessed. You persist in assuming
> that everyone who got the answer wrong guessed randomly, and there is no
> evidence of this.
>
Lookit, if they scored *exactly* the same as if they'd just guessed, then
ALL you have is proof that this is what they did. If they scored
significantly higher than the 3% standard error added onto the 25% they'd
get just for guessing [read: significantly higher than 28%], only then do
you have evidence that more than guessing was involved. If they scored 50%,
and the other 50% answered ONE of the wrong questions, then you may have
evidence that they didn't guess, but actually misunderstood the problem, or
were not taught properly. But if that 50% is spread evenly over the other
three answers, then you have statistical evidence that 50% of them GUESSED.
> >> This would only make sense if everyone in the group knew they didn't
> >> know, and just started guessing.
> >
> >Which was the case for many of the questions that American girls
answered.
> >You can't score that low, consistently, if you know something about the
> >subject
>> They did NOT score "that low, consistently"
>
Count the questions. Tell me what percentage of them scored lower than if
they'd just guessed http://christianparty.net/timssphysics.htm
>> >> My point is that you illustrate the saying, "Figures don't lie, but
> >> liars sure can figure."
> >>
> >> You figure because American girls got a score of 22.something correct
on
> >> one test question on one test, that there is no such thing as "gender
> >> equality".
> >
> >No. I figure because American girls consistently scored lower than if
> >they'd just guessed on a number of problems, not just one,
>> But they did not do so on all problems or even most, and indeed boys
scored
> low on nearly as many questions as girls.
>
Boys scored lower on exactly 2% of the questions. Girls scored lower than
boys on more than 90% of them.
> >and because this is a pattern that's repeated in many other tests,
>> Provide a citation to a real reference site (not one of yours) showing any
> data on any other standardized test question which is broken down by
gender
> showing or implying that girls guessed.
>> >that they were either
> >misinformed in the classroom, or didn't believe what they were taught and
> >went with "intuition" instead of facts (or a combination).
>> Or perhaps were not taught the relevant subject matter at all, and used
> imperfect logic to try to figure it out from what they did know.
>> >American boys didn't do that much better, btw, which is another thing
that I
> >figure.
>> As she said, liars figure.
>
The part that's always missing from that is "idiots can't detect when liars
are figuring". But you're right about one thing--an awful lot of American
girls and boys now fall into that category.
> >> > What part of that don't you understand (other than the typical and
> >> > infinitely STUPID statement by lojbab that no students guessed)?
> >>
> >> He never made that statement.
> >>
> >And I quote: "we don't know that people guess randomly when faced with a
> >test question they do not understand. Indeed, we know that they do not".
>> That is not the same thing as what I said. I said that *people* do not
guess
> *randomly*. "People" is not the same as "any person". It is quite
possible
> to find an occasional person who guessed randomly on a test - indeed *you*
> could prove that statement wrong by signing up for a test and then
guessing
> on all the answers, since it seems likely that you wouldn't understand any
of
> them. But in order to guess randomly, you would need to bring a random
> number generator in and use that to determine what answers you marked.
>> lojbab
You are real wierd. This was an EXACT quote of what you wrote, and I
predicted that you'd eventually deny you said it, but here you are denying
it the NEXT day.
Where in the world do you get the idea that a random number generator is
required for "any person" or *people* to make random choices?
It's impossible to imagine where you "liberals" get all these whacked out
notions. People and persons and a person make random choices all the time,
without computers, and with them. They don't need computers to make a
"random" choice on a multiple choice question, and MANY students do this
more often than they select the correct answer because they understand the
question.
John Knight
