No, since you average out the rank differences to calculate the correlation. The individual information is lost during the process.
[EMAIL PROTECTED] (CF) wrote in message news:<[EMAIL PROTECTED]>... > Hello, > > Given two sets (e.g., P and Q) of ordinal numbers associated with N variables, > I know how to calculate Spearman rank-order correlation coefficient > between them. > > Then, suppose what I know is > (i) the value of rank-order correlation coefficent between P and Q, > (ii) one set of ordinal numbers (e.g., P) > > Is there a way to estimate the other set (i.e., Q) of ordinal numbers > from (i) and (ii)? > > Thanks for your help, . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
