Hi, Sir Paul Yarnold, Your approach is really very good but this time I would like that you give an answer to my question. This time I writing results of a statistical study. In my statistical study, I have selected samples having a different number of paired data. Some of my samples have a small number of paired data and give a strong correlation of about 0.83. Thanks
No. The Pearson and Spearman correlation coefficients are measures of linear correlation not curvilinear correlation. Suppose there are a few data pairs. If the points are collinear, the correlation coefficients will be very high and precisely 1 or -1. Suppose the data pairs are now many. if they are not collinear, the correlation coefficients will be lower in magnitude. Hence r is not dependent on the number of data used.