Can somebody help me? It is pretty easy to obtain uncertainty of the slope (linear regression). But I don't know how to get the correct error if my data which i use for the linear regression have certain error themselves. Can somebody give me advice how to include the errors of individual points? To be more specific: from my initial data I obtained four linear regressions-As and their slopes and errors. those slopes I used as Y- data for second linear regression-B from which I need its slope and error. If I use the slopes (from As) just as four points my error is very high (25%) even though the correlation coefficient is 0.99. This is because I use T value for 95% confidence interval for n-2 points =2, which is very high. But as a matter of fact my initial data are based on a number of measurements (not just four), so I don't think my approach is correct. For chemist, maybe somebody can help me if I say that I'm calculating Activation energy and that first set of slopes are rate constants.
Thanks Alena ================================================================= Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =================================================================