Can we determine a correlation for what is not found, instead of what is found?

Thanks,

Yvonne

Printable View

- December 16th 2008, 06:17 PMyvonnehrcorrelation between what is not there?
Can we determine a correlation for what is not found, instead of what is found?

Thanks,

Yvonne - December 17th 2008, 10:58 AMmeymathis
Can you be more specific, or give an example? It just depends on what you mean.

If by correlate you mean compute a correlation coefficient, you can attempt to correlate any 2 variables in which you have paired numerical observations. The variables can even be categorical (0 means "ice", 1 means "Water", etc.), although the interpretation may be meaningless.