A question about correlation

Question

You So there's this assignment that requires you to work out the correlation between data that spans over a period of months and years. So for each month of each year there's a figure available. Anyway, so there's a set of data for one situation and one set for the other situation. I'm using excel to work out the correlation so would you advise that I use the correlation formula to work out the coefficient on a per month basis between these two sets of data and then take the average as my final answer? Or do I put all the data in at once, thus giving me the final answer?

Answer 1

The whole point of correlation is to discover how similar are TWO series. So, since you have two “situations”, and monthly data for each spanning several years, you definitely want to record the data for situation 1 into one column, say B, and the data for the other situation into another column, say C. You would put them in the same rows, corresponding to the same dates. (for example, put the data for January 2013 into B5 and C5, and put the data for February 2013 into B6 and C6, and so on)

Now the Excel CORREL function will give you the correlation between situation 1 and situation 2. Does that help?