[EMAIL PROTECTED] (Shi) wrote in message news:<[EMAIL PROTECTED]>... > > taking you at your word, i assume you really mean "perpendicular" and > > not vertical deviations. here's something that might help: > > > > www.mathpages.com/home/kmath110.htm > > Thanks : ) > > Yes I do mean "perpendicular" deviations. I had tried the method from > this website before, but for some reason that I didn't know, my > regression line was not exactly right ( -- by looking at the scatter > plot). The maximumlikehood estimator that I use now is from the book > "Analysis of Straight Line Data" by Forman S. Acton. It requires me to > have groups of data, and to know the correlation between the error > items before hand. However, all I have are two sets of measurements > that are taken by the same machine. I want to check the repeatibility > of the machine. What I thought of doing was to find the regression > parameters (the slope should be really close to 1) and their variance > so I could calculate the confidence intervals. If the parameters > estimated from my data set is outside of the 99% CI then I know there > is something going on with the machine since my last measurement. Any > suggestions on how to get the variance? > > Thanks to Diamond and anyone who look at this message : ) I really > appreciate your help. > > > Shi.
Ok, so you have the same units for both measures. now you want to check the repeatability of the machine. this sounds like agreement, rather than correlation. so with two sets of measures, would it make sense to calculate the intraclass correlation as a measure of agreement? . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
