Dear all, Yes, indeed my knowledge in statistics is rather limited, but let me reformulate my question. I have two samples (V1 and V2) of measurements of observed physical values( v1_i and v2_i). Each value in the samples is measured with an error (err_i), so it is in interval (v1_i-err_i, v1_i+err_i) for V1 sample, (v2_i-err_i, v2_i+err_i) for V2 sample. I need to know the probability of that these two samples have been drawn from the same distribution . I have already applied KS test (realized in CERN ROOT), but in the standard case it deals only with the values themselves (v1_i and v2_i) measuring the distance between their cumulatives. I would like to know if there a coded routine in any R package that could take into account errors (err_i) of EACH separate measurement (v1_i and v2_i) in my samples (V1 and V2). If there is no such routine, maybe you have an idea what algorithm (method) to apply for my problem. Thank you and regards, Igor. [[alternative HTML version deleted]]