I have several Pearson correlation coefficients with their corresponding p-values. But I need an Average Correlation Coefficient. To calculate it, I've used the Fisher z transformation mentioned here and here. But, probably due to my lack of knowledge, I didn't find a way to compute a p-value for this average correlation coefficient. Can you help me with this?
$\begingroup$
$\endgroup$
2
-
2$\begingroup$ What do you mean by an average correlation coefficient? Is this the mean of several correlations? $\endgroup$– Jeremy MilesCommented Apr 12, 2013 at 14:44
-
3$\begingroup$ Perhaps the answers at stats.stackexchange.com/questions/8019/… will be helpful. $\endgroup$– whuber ♦Commented Apr 12, 2013 at 17:32
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
You might just want to calculate the confidence interval.
$Z_r$ has a nearly normal distribution with variance
$$s_z^2 = 1 / n - 3.$$
Using these statistics we can construct a level $C$ confidence interval for the population value
$$Z_r \pm z^* / \sqrt{n - 3}$$
where $z^*$ is the critical value from the normal distribution such that the area between $-z^*$ and $z^*$ is equal to $C$.
