-1
$\begingroup$

Possible Duplicate:
Working with correlation coefficients

If I do Pearson correlation on Variable A (e.g. study hours) and Variable B (e.g. exam marks) for male students and female students, and I want to compare males with females, do I do a t-test or simply compare the correlation coefficients. Which is the better approach?

$\endgroup$
  • 2
    $\begingroup$ this is a good resource for this kind of questions. $\endgroup$ – mzuba Sep 16 '11 at 15:31
  • 1
    $\begingroup$ How would you even do a t-test on the two correlations? $\endgroup$ – Aniko Sep 16 '11 at 15:46
  • $\begingroup$ This question covers exactly the same ground as several recent questions. I recognize the slight variation introduced by the query about a t-test, but this difference is not great enough to warrant the explosion of new questions about what is essentially a single topic. $\endgroup$ – whuber Sep 16 '11 at 16:30
  • $\begingroup$ The "Working with correlation coefficients" link is not a possible duplicate! $\endgroup$ – Adhesh Josh Sep 17 '11 at 11:10
0
$\begingroup$

This all depends on what question you are trying to answer. Do you want to find out if exam marks are different on averege for males and females? Do you want to find out if the relationship between study hours and marks is different between males and females? Do you want to find out if marks are different on average after adjusting for possible differences in study hours? or something else?

The first task is to figure out what your question is.

$\endgroup$
  • $\begingroup$ Thanks. I am trying to find if the relationship between study hours and marks are different between males and females. $\endgroup$ – Adhesh Josh Sep 17 '11 at 10:56

Not the answer you're looking for? Browse other questions tagged or ask your own question.