"Parse Tree" <parsetree at hotmail.com> wrote in message
news:V1B%8.10383$sb5.1039243 at news20.bellglobal.com...
> "John Knight" <johnknight at usa.com> wrote in message
> news:XHy%8.17561$Fq6.2085937 at news2.west.cox.net...> >
> > "Parse Tree" <parsetree at hotmail.com> wrote in message news:d31%8.8024
> > > > So when 50% get the answer correct on a four part multiple choice
> > > question,
> > > > and if students just guessed at the answers they didn't know or
> > understand
> > > > (which is the worst case scenario), then (ignoring the standard
error)
> > as
> > > > little as one third of them actually demonstrated a knowledge or
> > > > understanding of the question.
> > > >
> > > > Let's look at what happens if 90% got it correct. x = total
guesses,
> > .25x
> > > =
> > > > correct guesses, .75x = incorrect guesses = 10 percent, x = 13.33
> > percent,
> > > > .25x = 3.33 percent = correct guesses, 90% got it correct - 3.33%
got
> it
> > > > correct by guessing = 86.67% demonstrated a knowledge or
understanding
> > of
> > > > the question.
> > >
> > > How are you calculating the number of people that guessed? I would
like
> > to
> > > see how this is possible, since I know it is not.
> > >
> > >
> >
> > Let's pretend that you really are asking a serious question, that you
> really
> > want to know, that this isn't intended to be a feminazi diversionary
> tactic,
> > that you really couldn't figure it out from prior posts, and that your
> > question deserves a serious answer.
> >
> > If you're faced with a question that you have utterly no idea even what
> the
> > words mean, that couldn't be more confusing to you if it were written in
> > Greek (which the 0% correct response rate of American girls on many of
the
> > non-multiple choice questions demonstrated was the case), and you have
> four
> > choices on a multiple choice question--what do you do?
> >
> > Do you ask a neighbor? Do you call someone on your cell phone? Do you
> ask
> > God for the answer?
> >
> > Or do you guess?
> >
> >
> > ps--here's a hint: the only time that getting a 25% correct response
rate
> > on a four choice multiple choice question would not represent 100%
guesses
> > is when a majority answered the wrong question, which *might* be
evidence
> > that they were taught the wrong thing instead.
>> Oh, then you really are categorically wrong. You have no method of
> calculating the number of guesses, and you work based on a number of
> incorrect assumptions.
>>
First of all, it's not an incorrect assumption.
Second of all, it's not categorically wrong.
Thirdly, the proper procedure if you want to critique the method is to
provide specifically what you "think" is an incorrect assumption or is
wrong.
Fourthly, the least you could do is provide at least a template for what you
believe to be the correct way to do it, if you *really* think this is
incorrect.
John Knight
