Math people...EASY statistical analysis question

Open discussion about any topic, as long as you abide by the rules of course!
Post Reply
tnf
Posts: 13010
Joined: Tue Mar 13, 2001 8:00 am

Math people...EASY statistical analysis question

Post by tnf »

For some reason my brain has froze as I am trying to put some data into a spreadsheet for analysis.

This is ridiculously easy, but for some reason I don't think I'm doing it right...kind of like when you suddenly can't spell a word you use all the time.

Here's the deal.
I've got a test I give that has each question scored on a 4,3,2,1 point scale. I need to figure out the average score per question based on data like this:

Let's say I have 43 total tests I am collecting data for.

Question 1 had 21 four point answers, 10 three point answers, 7 two point answers and 5 one point answers. I need to figure out what the value for the average score was (i.e. 3.2, 2.1, or whatever...) Someone set this up for me please.

I've got a formula in excel that I think is correct, but I'm not certain.
mjrpes
Posts: 4980
Joined: Tue Nov 28, 2000 8:00 am

Post by mjrpes »

( (21 * 4) + (10 *3) + (7 * 2) + (5 * 1) ) / (21 + 10 + 7 + 5)
tnf
Posts: 13010
Joined: Tue Mar 13, 2001 8:00 am

Post by tnf »

eye, that's what I had. But for some reason it just didn't seem right.
thanks.
mjrpes
Posts: 4980
Joined: Tue Nov 28, 2000 8:00 am

Post by mjrpes »

can't see how it would be different.

it might seem wrong because we are dropping out additional steps. the actual equation is

total points achieved / total points possible = percentage correct

this would be

((21 * 4) + (10 *3) + (7 * 2) + (5 * 1)) / ((21 + 10 + 7 + 5) *4))

but this is only the percentage correct. since we want to total correct out of 4, we times this whole thing by 4.

((21 * 4) + (10 *3) + (7 * 2) + (5 * 1)) / ((21 + 10 + 7 + 5) *4)) * 4

but the 4's can cancel out at this point.

mebbe that's why :paranoid:
Nightshade
Posts: 17020
Joined: Fri Dec 01, 2000 8:00 am

Post by Nightshade »

Weighted averages?
menkent
Posts: 2629
Joined: Sun Jul 23, 2000 7:00 am

Post by menkent »

mjrpes wrote:( (21 * 4) + (10 *3) + (7 * 2) + (5 * 1) ) / (21 + 10 + 7 + 5)
3.1 seems about right given the distribution :shrug:
Post Reply