I just tried to do a correlation with the china study data from online. I took the Blood plasma retinol levels against the retinol and carotenoid without correcting for anything, and found that for the same locations, the ones that consumed the most vit. A had the lowest blood levels. The actual level is so small as to be statistically insignificant, but what am I doing wrong? I take the data that the file says marks the different values, take the columns, and plot them versus the other one. This is why I hate correlational studies. Unfortunately MDA doesn't let us post up excel files as attachments. Note that I'm using the excel linear method which I think is least squares. I plan to check Minger's work on atleast one correlation just to be sure.