Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.



Hypothesis test for correlation formula
Posted:
Jul 25, 2006 1:42 PM


Hi  I hope this question makes sense. I'm trying to write a little computer program dealing with correlation, and I'd like to be able to test whether the correlation coefficient (r) falls below a critical level. All the books I have on this have a table in the back of the book that usually have a heading along the lines of "Critical Levels of r at 5% and 1% Levels of Significance." So the critical level at 5%, with 4 degrees of freedom, would be .811. The critical level at 1%, with 6 degrees of freedom, is .834, and so on.
Does anyone know how the numbers in this table are generated? I'd like to be able to write a little computer routine, where I enter the degrees of freedom and percent level, and calculate the critical level. I'm looking for a formula that I can use. Any help would be appreciated.
Thanks, Steve Sinai



