Hello, everyone. We have some regression data we'd like to report, but aren't sure how to best do so. Here's the situation:

A lab analyst runs a single experiment at a given setting which produces a response. he then performs additional experiments at different settings and collects the responses.

At this point, he fits a regression line to the data. To an analytical chemist, this is called a calibration curve--it can be given to another analyst who can use it to predict a response at a given experimental setting.

Say the curve is given to an analyst who wants to report the predicted value from the regression equation at a given setting....what's the best way to report the uncertainty of the predicted value?

Thank you.

Artie