Hi everyone, In social reseach, economic and others fields "Non-Equivalent Groups Analysis" shows an issue when data are biased between Posttest and Pretest. Many authors talk about the Regression-Discontinuity Analysis wich is used when dispersion are different in both groups (in this case Control vs. Treatment). I have found this function that seems to work well finding this problem (sm.discontinuity), also fixing this problem in is not trivial among reseachers. A good source of theorical information correcting this problem can be found here: http://www.socialresearchmethods.net/kb/statnegd.php Someone knows if there's any way to do this through R? Thanks in advance! José [[alternative HTML version deleted]]