Struggling with written statistics question.

Frogs

New Member
#1
Hello.
I'm currently revising for a statistics re-sit examination and when going over past papers have found one question commonly appears.
Both Pearson's correlation and a regression table are given, with the pearsons correlation between two variables being significant, but the beta regression for the same two variables not being significant.
The questions asks for possible reasons why the correlation may be statistically significant at the 0.05 significane level, but the regression beta is not.
I've searched through all the course books but can't find any information suggesting why this is, when they are both comparisons of the same relationships.
Any help or pointers, even ideas, would be very much appreciated.
Thank you.
Andy.
 

JohnM

TS Contributor
#2
I don't think it's possible. The significance of both R and beta are determined by using the same t statistic.

Trick question?