Measure Aggregation - Sum or Average? How do I know?

[yac431]

Hi Data Experts, 

I am writing to ask, while working on a data set the measures are always aggregated as Sum as a default. What should we consider if we want to make the Aggregation as Average. And why? I am creating a correlation matrix for all independant variables in a Housing Data set. The independant variables are Characteristics of the houses. Example - no. of bathrooms, Plot size Sqft, Bedrooms Above Grade, Etc. My next task is to create a cluster analysis out of these variables and then match one cluster ID with the Dependent variable. I.e. the Sales prices of the Houses. The main purpose of this is to do a cluster analysis of the data advising the client to gain insights on the characteristics of different types of houses Sold. 

Should I be taking the Average of these measures to make a correlation matrix? 

[PaigeMiller]

Correlations are not done on sums, correlations are not done on averages. Correlations are computed from individual (un-summed, un-averaged) data. So, I don't understand the question.

[Stu_SAS]

Hey @yac431! @PaigeMiller is correct. For this particular object, aggregations do not affect it. The correlations are derived on un-summed data based on the Pearson Product Moment Correlation Coefficient. This formula is: