correlation question

#1
I have a question regarding a simple correlation of estimated rather than known values.

Let’s say I am interested in the correlation between A and B. Values for A and B were estimated instead of known (they have associated sampling error). Does this uncertainty influence the correlation? If so, how can I adjust it to account for the error of estimation?

Thanks
 

JohnM

TS Contributor
#2
I don't believe the uncertainty affects the correlation. I've done similar correlations and I don't recall making any adjustments.

Any set of measurements you take is a random sample of all of the possible measurements, so any parameters derived from them are estimates as well.