"mat" <mats_trash at hotmail.com> wrote in message
news:43525ce3.0207260248.ef11825 at posting.google.com...
> "John Knight" <johnknight at usa.com> wrote in message
news:<_2W%8.20049$Fq6.2409696 at news2.west.cox.net>...
> > "mat" <mats_trash at hotmail.com> wrote in message > > So let's make it
REAL
> > simple. Let's define a simple statement of the
> > > > problem.
> > > >
> > > > If you have 100,000 students *randomly guessing* at one multiple
choice
> > > > question which has four possible answers (A., B., C., and D.), one
of
> > which
> > > > must be selected, then there is only ONE possible outcome:
> > > >
> > > > A. gets selected 25,000 times.
> > > >
> > > > B. gets selected 25,000 times.
> > > >
> > > > C. gets selected 25,000 times.
> > > >
> > > > D. gets selected 25,000 times.
> > > >
> > >
> > > Your contradicting yourself and you don't even realise it. Given that
> > > the students are guessing 'randomly' then it is by no means certain
> > > that anything of the sort you describe is going to happen (thats what
> > > random means). What probability tells you is what is more or less
> > > likely to happen and what would happen as sample size and repetition
> > > tends toward infinity
> > >
> >
> > Mat, this is getting to be a waste of time. You're actually going
> > BACKWARDS, adding to the negative knowledge, and confusing more people
than
> > just yourself.
> >
> > If you don't believe the above is precisely what will happen (plus or
minus
> > 0.75%), then exactly what do you *think* will happen. Be specific, and
tell
> > us exactly why you *think* that.
> >
> > This is fundamental to getting the rest correct, so please don't go
blasting
> > off into territory that is clearly going to confuse the issue, and focus
> > only on the above.
> >
> > Tell us why the above won't be the outcome, and tell us what you *think*
it
> > will be instead.
> >
> > John Knight
>> Don't you get it yet - I'm not saying the above won't be the outcome
> but the whole basis of probability is that it tells you the relative
> likelyhood of events occuring, not what IS going to happen, (unless
> the probability is 1, in which case the event is certain) . You've
> confused yourself by working out that answering one question correct
> is certain thus the distribution of answers must be as above. ITS
> WRONG. plain as that. What I think will occur is that if everyone
> randomly guesses then the result may well be something like an even
> distribution across each answer but it may well be that many more
> people choose a certain answer. In short, I don't know though I can
> make an estimate given the probabilties of different outcomes.
>> Let me ask you this - what is the probability of getting one answer
> correct having answered five questions? is it 1.25 (as it would be by
> your logic)? Whats your answer?
>> Why didn't you partake in the little test I constructed? The answers
> are already on the post (not explicitly), so I can't cheat.
In a backhanded way, you just acknowledged that the above distribution IS
correct, and that you CANNOT present any FACTS to dispute it, so at least
thanks for that.
You're making a common and serious mistake regarding your application of
probability theory, and it's not even worth beginning to explain it to you.
It's easy enough to work out the math--BUT THAT'S NOT THE POINT.
Once you correct the following error, you're home free, without having to
run around in left field for the rest of the day:
> What I think will occur is that if everyone
> randomly guesses then the result may well be something like an even
> distribution across each answer but it may well be that many more
> people choose a certain answer.
Exactly who do you think *cares* what you "think"?
What you claim you "think" is precisely the OPPOSITE of the FACTS, so what
you "think" is less than irrelevant.
By DEFINITION, if the selection *is* random, then this WILL be the outcome:
end of story.
There are pretty little pictures at
http://christianparty.net/randomdistribution.htm which might help you
visualize the problem [that is, IF feminazis are really capable of grasping
what graphs like this mean, which I'm beginning to doubt].
In your defense, the ACTUAL outcome on many of the TIMSS questions did NOT
exhibit a random selection. Certainly when the responses were lower than if
the girls had just guessed on ONE THIRD of the questions, this isn't a
random selection--BUT THE POINT IS THAT WE DO KNOW WHAT THE OUTCOME WOULD BE
if IT *had* BEEN A RANDOM SELECTION, which can be compared to the actual
outcome, which enables us to evaluate the problem (which was the whole
purpose of TIMSS in the first place), which will go a LONG way toward
finding a solution, which will ultimately benefit even the feminazis who
want nothing but to shoot the messenger.
John Knight
