Skip to main content
Cross Validated

Return to Question

deleted 6 characters in body
Source Link
Alexis
Alexis
  • 30.7k
  • 8
  • 101
  • 176

I am writing a meta-analysis looking at the correlation between different tests for assessing body composition.

The results of the studies that I have included use different ways to calculate the correlation, including linear regression (r squared$R^{2}$), pearson's rPearson's $r$, concordance correlation coefficient (CCC), and mean difference (SD) + 95% CI.

I would like to compare these results. I have found that I can take the square root of r squared$\sqrt{R^{2}}$ to get r$r$. Is there also a way to calculate r$r$ from CCC? Or is CCC so similar to pearson's rPearson's $r$, that I can just call it r$r$ and throw everything on one pile?

I am writing a meta-analysis looking at the correlation between different tests for assessing body composition.

The results of the studies that I have included use different ways to calculate the correlation, including linear regression (r squared), pearson's r, concordance correlation coefficient (CCC), and mean difference (SD) + 95% CI.

I would like to compare these results. I have found that I can take the square root of r squared to get r. Is there also a way to calculate r from CCC? Or is CCC so similar to pearson's r, that I can just call it r and throw everything on one pile?

I am writing a meta-analysis looking at the correlation between different tests for assessing body composition.

The results of the studies that I have included use different ways to calculate the correlation, including linear regression ($R^{2}$), Pearson's $r$, concordance correlation coefficient (CCC), and mean difference (SD) + 95% CI.

I would like to compare these results. I have found that I can take $\sqrt{R^{2}}$ to get $r$. Is there also a way to calculate $r$ from CCC? Or is CCC so similar to Pearson's $r$, that I can just call it $r$ and throw everything on one pile?

edited tags
Link
Ferdi
Ferdi
  • 5.3k
  • 10
  • 47
  • 64
Source Link
Nicole IJff
Nicole IJff
  • 51
  • 1
  • 2

How to compare concordance correlation coefficient to Pearson's r?

I am writing a meta-analysis looking at the correlation between different tests for assessing body composition.

The results of the studies that I have included use different ways to calculate the correlation, including linear regression (r squared), pearson's r, concordance correlation coefficient (CCC), and mean difference (SD) + 95% CI.

I would like to compare these results. I have found that I can take the square root of r squared to get r. Is there also a way to calculate r from CCC? Or is CCC so similar to pearson's r, that I can just call it r and throw everything on one pile?