Correlation Coefficient

[M.Goetz]

Have utilized my Weibull++6 software and have gotten the Distribution Wizard to determine which distribution best fits the data. If I compare the Correlation Coefficient between two distributions (Rank Regression method) it shows that my 2nd distribution is slightly better at fitting the linear fit? Any easy way of understand this? thanks

[RS Support]

I guess I am little bit confused. Is the question about the corr. coef. (cc) or about the wizard. I’ll try to answer both.

The corr. coef. is a measure of the linearity of the data on the appropriate plotting paper. Now the paper is constructed so that it linearizes the cdf of the distribution. Thusm if the data is linear on the paper the more valid is the assumption that the data comes from that distribution.

The wizard uses more than one metric (one of which is the corr. coef.).