vpFREE2 Forums

XVP: Need Help From the Math Experts

I'm writing a quiz, and I'm having a problem with the scoring. I hope you can help.

The quiz consists of five questions, and each question has four possible answers. Each of the questions are structured so that every "A" answer fits one profile, every "B" answer fits another profile, every "C" answer fits a third profile, and every "D" answer fits a fourth profile.

Without asking respondants to compute averages and keeping the scoring as simple as possible, how do you score the quiz so that the overall numerical score can be used to assign generic profiles to respondants and answer the question, "If you scored ___, this means..."

Thanks for your help!
Lainie

···

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

Yeah sure, I'm a math expert. If you can't find an answer here then go to the statistical forum on twoplustwo.com Those guys fire real bullets from trenches.
JT

···

Lainie Wolf <LainieWolf@hotmail.com> wrote:
I'm writing a quiz, and I'm having a problem with the scoring. I hope you
can help.

The quiz consists of five questions, and each question has four possible
answers. Each of the questions are structured so that every "A" answer fits
one profile, every "B" answer fits another profile, every "C" answer fits a
third profile, and every "D" answer fits a fourth profile.

Without asking respondants to compute averages and keeping the scoring as
simple as possible, how do you score the quiz so that the overall numerical
score can be used to assign generic profiles to respondants and answer the
question, "If you scored ___, this means..."

Thanks for your help!
Lainie

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

vpFREE Links: http://members.cox.net/vpfree/Links.htm

---------------------------------
YAHOO! GROUPS LINKS

    Visit your group "vpFREE" on the web.
  
    To unsubscribe from this group, send an email to:
vpFREE-unsubscribe@yahoogroups.com
  
    Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

---------------------------------

---------------------------------
Yahoo! Music Unlimited - Access over 1 million songs. Try it free.

[Non-text portions of this message have been removed]

Thanks! It seems like it will take a day or so to complete my
registration at that site; in the meantime, I'm hoping that I'll find
some answers here...

Lainie

--- In vpFREE@yahoogroups.com, JT Hughes_iii <jt_hughes_iii@y...>
wrote:

Yeah sure, I'm a math expert. If you can't find an answer here then

go to the statistical forum on twoplustwo.com Those guys fire real
bullets from trenches.

Not a math expert either, but all you need is an orthogonal encoding
scheme. A particullary simple one is to use arbitray base number
system, for example:
A =10^0 = 1
B= 10^1 = 10
C=10^2 = 100
D= 10^3 = 1000

Not pretty, but it works. You can easily use any base >= 5 (and many
different series of exponents).

The quiz consists of five questions, and each question has four

possible answers. > Thanks for your help!

Lainie

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today -

it's FREE!

···

--- In vpFREE@yahoogroups.com, "Lainie Wolf" <LainieWolf@h...> wrote:

http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

Thanks, but I'm not sure I understand this model enough to set the "if you score is in this range, this means..." (For the purposes of the rest of this email, I'll call the ranges "buckets.")

Here's what I'm finding confusing: I''d expect the "A" bucket to have a low numerical range (starting with a score of 5); the "B" bucket to have a higher numerical range; the "C" bucket to have still a higher range; and the "D" bucket to have the highest range (topping out at 5000).

If you answered all the questions with an "A," your score would be 5. If you answered all the questions with a "D," your score would be 5000.

Logically, since there are five questions, if at least three of your answers were the same, that 'majority answer' should dictate the bucket you get placed into.

But what happens when the answers vary?

If you answers were 3 As and 2 Ds, your score would be 2003 (and you'd probably wind up in the "C" or "D" bucket). If you scored 3 Bs and 2 As, your score would be 32 (and you'd probably wind up in the "A" bucket).

What am I missing?

Thanks again for your help.

Lainie

Not a math expert either, but all you need is an orthogonal encoding scheme. A particullary simple one is to use arbitray base number system, for example:
A =10^0 = 1
B= 10^1 = 10
C=10^2 = 100
D= 10^3 = 1000

Not pretty, but it works. You can easily use any base >= 5 (and many
different series of exponents).

The quiz consists of five questions, and each question has four

possible answers. > Thanks for your help!

···

cdfsrule <groups.yahoo@verizon.net> wrote:
--- In vpFREE@yahoogroups.com, "Lainie Wolf" <LainieWolf@h...> wrote:

Lainie

---------------------------------
Yahoo! Music Unlimited - Access over 1 million songs. Try it free.

[Non-text portions of this message have been removed]

You're not really missing anything. Generally speaking, when you
take a 'quiz' in a reader's digest or the like, they are generally
coded so that A = 1, B = 2, C = 3, D = 4. The more you spread out
the values, the easier it is to stratify the responses, but the
harder some poor shnook who is self marking himself has to work.

Generally speaking, the answers should show a spectrum of
the 'trait' you're looking for, ranging from A to D.

So, if a person answers with 3 A's and 2 B's, they are an A/B person.

If they answer with 1 A, 2B, 1C, 1D, then who the heck knows what
they are, and they will wind up between B & C, which is about as
accurate as you're going to get with 5 questions.

So, if you use the scoring above, you'll know that a '5' in total is
a solid A and a 20 is a solid D. a 19 can only be achieved with
4 'D' answers and 1 'C', so the person is still essentially a D.

If you map out the possibilities, you'll begin to realize that 18-20
is a 'D' type, 13-17 is a C type, 8-12 is a B type. 5-7 is an A.
Obviously, the numbers closest to the edge will hardly be a
scientific judgemnt.

You can chose to use 10, 100, 1000, 10000 and it will become a bit
more obvious where the person lies because you will know EXACTLY how
many A's, B's, C's and D's the person answered and you can decide
exactly what overall mark to give them.

Just a few thoughts that I hope help.

Kil

--- In vpFREE@yahoogroups.com, Lainie Wolf <lainiewolf702@y...>
wrote:

Thanks, but I'm not sure I understand this model enough to set

the "if you score is in this range, this means..." (For the
purposes of the rest of this email, I'll call the ranges "buckets.")

Here's what I'm finding confusing: I''d expect the "A" bucket to

have a low numerical range (starting with a score of 5); the "B"
bucket to have a higher numerical range; the "C" bucket to have
still a higher range; and the "D" bucket to have the highest range
(topping out at 5000).

If you answered all the questions with an "A," your score would be

5. If you answered all the questions with a "D," your score would
be 5000.

Logically, since there are five questions, if at least three of

your answers were the same, that 'majority answer' should dictate
the bucket you get placed into.

But what happens when the answers vary?

If you answers were 3 As and 2 Ds, your score would be 2003 (and

you'd probably wind up in the "C" or "D" bucket). If you scored 3
Bs and 2 As, your score would be 32 (and you'd probably wind up in
the "A" bucket).

What am I missing?

Thanks again for your help.

Lainie

cdfsrule <groups.yahoo@v...> wrote:
Not a math expert either, but all you need is an orthogonal

encoding scheme. A particullary simple one is to use arbitray base
number system, for example:

A =10^0 = 1
B= 10^1 = 10
C=10^2 = 100
D= 10^3 = 1000

Not pretty, but it works. You can easily use any base >= 5 (and

many

different series of exponents).

--- In vpFREE@yahoogroups.com, "Lainie Wolf" <LainieWolf@h...>

wrote:

···

> The quiz consists of five questions, and each question has four
possible answers. > Thanks for your help!
> Lainie
>

---------------------------------
Yahoo! Music Unlimited - Access over 1 million songs. Try it free.

[Non-text portions of this message have been removed]

Thank you very much. This is extremely helpful!

Best,
Lainie

You're not really missing anything. Generally speaking, when you take a 'quiz' in a reader's digest or the like, they are generally coded so that A = 1, B = 2, C = 3, D = 4. The more you spread out the values, the easier it is to stratify the responses, but the harder some poor shnook who is self marking himself has to work.

Generally speaking, the answers should show a spectrum of
the 'trait' you're looking for, ranging from A to D.

So, if a person answers with 3 A's and 2 B's, they are an A/B person.

If they answer with 1 A, 2B, 1C, 1D, then who the heck knows what they are, and they will wind up between B & C, which is about as accurate as you're going to get with 5 questions.

So, if you use the scoring above, you'll know that a '5' in total is a solid A and a 20 is a solid D. a 19 can only be achieved with 4 'D' answers and 1 'C', so the person is still essentially a D.

If you map out the possibilities, you'll begin to realize that 18-20 is a 'D' type, 13-17 is a C type, 8-12 is a B type. 5-7 is an A. Obviously, the numbers closest to the edge will hardly be a scientific judgemnt.

You can chose to use 10, 100, 1000, 10000 and it will become a bit more obvious where the person lies because you will know EXACTLY how many A's, B's, C's and D's the person answered and you can decide exactly what overall mark to give them.

Just a few thoughts that I hope help.

Kil

--- In vpFREE@yahoogroups.com, Lainie Wolf <lainiewolf702@y...>
wrote:

Thanks, but I'm not sure I understand this model enough to set

the "if you score is in this range, this means..." (For the
purposes of the rest of this email, I'll call the ranges "buckets.")

Here's what I'm finding confusing: I''d expect the "A" bucket to

have a low numerical range (starting with a score of 5); the "B"
bucket to have a higher numerical range; the "C" bucket to have
still a higher range; and the "D" bucket to have the highest range
(topping out at 5000).

If you answered all the questions with an "A," your score would be

5. If you answered all the questions with a "D," your score would
be 5000.

Logically, since there are five questions, if at least three of

your answers were the same, that 'majority answer' should dictate
the bucket you get placed into.

But what happens when the answers vary?

If you answers were 3 As and 2 Ds, your score would be 2003 (and

you'd probably wind up in the "C" or "D" bucket). If you scored 3
Bs and 2 As, your score would be 32 (and you'd probably wind up in
the "A" bucket).

What am I missing?

Thanks again for your help.

Lainie

cdfsrule <groups.yahoo@v...> wrote:
Not a math expert either, but all you need is an orthogonal

encoding scheme. A particullary simple one is to use arbitray base
number system, for example:

A =10^0 = 1
B= 10^1 = 10
C=10^2 = 100
D= 10^3 = 1000

Not pretty, but it works. You can easily use any base >= 5 (and

many

different series of exponents).

--- In vpFREE@yahoogroups.com, "Lainie Wolf" <LainieWolf@h...>

wrote:

> The quiz consists of five questions, and each question has four
possible answers. > Thanks for your help!
> Lainie
>

---------------------------------
Yahoo! Music Unlimited - Access over 1 million songs. Try it free.

[Non-text portions of this message have been removed]

vpFREE Links: http://members.cox.net/vpfree/Links.htm

SPONSORED LINKS
Gambling Online gambling Outdoor recreation Recreation software

···

kilroydq2 <kilroydq@prodigy.net> wrote:

---------------------------------
YAHOO! GROUPS LINKS

    Visit your group "vpFREE" on the web.
  
    To unsubscribe from this group, send an email to:
vpFREE-unsubscribe@yahoogroups.com
  
    Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

---------------------------------

---------------------------------
Yahoo! Music Unlimited - Access over 1 million songs. Try it free.

[Non-text portions of this message have been removed]