I am working with a dataset and I find that a polynomial trend line gives a better r^2 value. Is Pearson\'s R/Pearson\'s Correlation Coefficient still a good indicator in th