My brain has turned to mush after spending too much time on the internet. I'm trying to remember a math-formula I used to use but I can't remember what it was. It looks like Linear Regression but there was another component to it whereby a probability of correlation was calculated as it approached either 1 or -1. For example, 0.7 wasn't so probable but 0.99 or -0.98 was very probable. I also seem to remember it looked like the Linear Regression formula BUT I think it had a big Square Root over it. Anyone got any clue what the hell I'm talking about? It's driving me nuts trying to remember.
Aha! Got it! It's the correlation coefficient I was trying to name. http://en.wikipedia.org/wiki/Correlation_coefficient
Hi ... were you looking for Pearson's or Spearman's? Pearson for interval and ratio data; Spearman for ordinal. Makes a big difference, although you can axiomatically assume ordinal = interval, if a zero is theoretically possible. Good luck.
Thanks but I can't actually remember! It's amazing how much you forget all this stuff that used to be second nature.