Thread: Correlation Coefficients

Hey everyone, I was wondering if you can help me conceptualize something.

I was running a few correlations in excel this morning and one of them returned a coefficient of 19.38. Now, I obviously have no formal education in statistics, but I was under the impression that as the coefficient approaches 1.0, the 'stronger' the relationship. So what does it mean when the coefficient is above 1?

So, I just noticed both sets of values are percentages. Perhaps excel is converting the coefficient (.1938) to a percent (19.38) for me? That makes more sense, correct?

Originally Posted by JDA

So, I just noticed both sets of values are percentages. Perhaps excel is converting the coefficient (.1938) to a percent (19.38) for me? That makes more sense, correct?

Although I don't know much of excel but I guess its in percentage only...

ok, new question:

I changed all the percentages to decimals (i.e. instead of 20% each cell now reads .20) and the coefficient for the same data sets now reads .164? How is the relationship now weaker? Why has it changed?

Originally Posted by JDA

ok, new question:

I changed all the percentages to decimals (i.e. instead of 20% each cell now reads .20) and the coefficient for the same data sets now reads .164? How is the relationship now weaker? Why has it changed?

I guess this is absolutely related to excel only...not sure of it!!