View Single Post
Old 10-01-2012, 03:16 PM   #4 (permalink)
Moderating From Above
Harbinger[TG]'s Avatar
Join Date: Apr 2007
Location: Riding A Pale Horse
Send a message via MSN to Harbinger[TG]

Fan of EMR
Originally Posted by StJimmy666 View Post
Alright, debating with someone and I need a math person:

Given 2 sets of data (10 values each) so X and Y, how would you solve for the following:

(Summation Sign)XY

It's the sum of X times Y right? So how would you solve for it?
If you are only referring to this "(Summation Sign)XY" then it should be the XY1+XY2+ etc etc. There is actually this What does Σxy mean (sigma xy)? - Yahoo! Answers

Originally Posted by super_gman View Post
The way you have that equation written out, I would read it as SUM(X*Y). So you would multiply all the X's times their corresponding Y's, and add up all the products. Or, for the visually inclined:

X1Y1 + X2Y2 + X3Y3 + X4Y4 + X5Y5 + X6Y6 + X7Y7 + X8Y8 + X9Y9+ X10Y10

Of course, I say that under the assumption that the scores are somehow corresponding to one another, such as in a correlation. The original post wasn't 100% clear on that point (or perhaps I'm just not reading it properly).
I agree. I don't really see why one would be calculating this because the two data sets have nothing to do with another.

Originally Posted by StJimmy666 View Post

So my TA is indeed a dingbat and I'm not silly. This is great news.

The scenario is: The X set is the values of cigarettes smoked by 10 Americans daily, Y is the values of cigarettes smoked by 10 Europeans daily.

The questions were

sumXY (Sum of X times Y)

sumXsumY (Sum of X times Sum of Y)

sumXY - sumXsumY/N over N-1 (I hate math)
Yeah. i don't see how the two data sets are related and would require this. Unless it is just to do it for maths sake. Might want to also check this out Statistics

Another statistic is the covariance sigmaxy defined by:

sigmaxy = [(x1 - mux)(y1 - muy) + ... + (xn - mux)(yn - muy)]/n = (x1y1 + ... + xnyn)/n - muxmuy.

This is a measure of how much changes in x are associated with changes in y. If sigmaxy > 0 then x and y tend to increase or decrease together. If sigmaxy < 0, then y tends to decrease as x increases, and vice versa. If sigmaxy = 0, or is very small, then x and y tend to be independent of each other.

What class is this for? been awhile since I had to do any of this.

Might want to watch this How To Calculate Covariance (Math)

I made a quick example in excel looking for the 3 explanations. I think this is accurate.

1 2 2
2 2 4
3 2 6
4 2 8
5 2 10
6 2 12
7 2 14
8 2 16
9 2 18

sumx 45
sumy 18
sumXY 90

sumXY/n 10
sumXY/n/n-1 1.25

Last edited by Harbinger[TG]; 10-01-2012 at 03:25 PM.
Harbinger[TG] is offline   Reply With Quote