Concordance correlation coefficient

From Wikipedia, the free encyclopedia
Jump to: navigation, search

In statistics, the concordance correlation coefficient measures the agreement between two variables, e.g., to evaluate reproducibility or for inter-rater reliability.

Definition[edit]

Lawrence Lin has the form of the concordance correlation coefficient \rho_c as[1]

\rho_c = \frac{2\rho\sigma_x\sigma_y}{\sigma_x^2 + \sigma_y^2 + (\mu_x - \mu_y)^2},

where \mu_x and \mu_y are the means for the two variables and \sigma^2_x and \sigma^2_y are the corresponding variances. \rho is the correlation coefficient between the two variables.

This follows from its definition[1] as

\rho_c = 1 - \frac{{\rm Expected\ orthogonal\ squared\ distance\ from\ the\ diagonal\ }x=y}
{{\rm Expected\ orthogonal\ squared\ distance\ from\ the\ diagonal\ }x=y{\rm \ assuming\ independence}}.

When the concordance correlation coefficient is computed on a N-length data set (i.e., two vectors of length N) the form is

\hat{\rho}_c = \frac{2 s_{xy}}{s_x^2 + s_y^2 + (\bar{x} - \bar{y})^2},

where the mean is computed as

\bar{x} = \frac{1}{N} \sum_{n=1}^N x_n

and the variance

s_x^2 = \frac{1}{N} \sum_{n=1}^N (x_n - \bar{x})^2

and the covariance

s_{xy} = \frac{1}{N} \sum_{n=1}^N (x_n - \bar{x})(y_n - \bar{y}) .

Whereas the ordinary correlation coefficient (Pearson's) is immune to whether the biased or unbiased versions for estimation of the variance is used, the concordance correlation coefficient is not. In the original article Lin suggested the 1/N normalization,[1] while in another article Nickerson appears to have used the 1/(N-1),[2] i.e., the concordance correlation coefficient may be computed slightly differently between implementations.

Relation to other measures of correlation[edit]

The concordance correlation coefficient is nearly identical to some of the measures called intra-class correlations, and comparisons of the concordance correlation coefficient with an "ordinary" intraclass correlation on different data sets found only small differences between the two correlations, in one case on the third decimal.[2] It has also been stated[3] that the ideas for concordance correlation coefficient "are quite similar to results already published by Krippendorff[4] in 1970".

In the original article[1] Lin suggested a form for multiple classes (not just 2). Over ten years later a correction to this form was issued.[5]

One example of the use of the concordance correlation coefficient is in a comparison of analysis method for functional magnetic resonance imaging brain scans.[6]

External links[edit]

  • Statistical Calculator. Provided by NIWA, it is an online version of Lin’s concordance used to assess the degree of agreement between two continuous variables, such as chemical or microbiological concentrations. It calculates the value of Lin’s concordance correlation coefficient. Values of ±1 denote perfect concordance and discordance; a value of zero denotes its complete absence. Statistical testing procedures for Cohen's kappa and for Lin’s concordance correlation coefficient are included in the calculator. These procedures guard against the risk of claiming good agreement when that has happened merely by "good luck".

References[edit]

  1. ^ a b c d Lawrence I-Kuei Lin (March 1989). "A concordance correlation coefficient to evaluate reproducibility". Biometrics (International Biometric Society) 45 (1): 255–268. doi:10.2307/2532051. JSTOR 2532051. PMID 2720055. 
  2. ^ a b Carol A. E. Nickerson (December 1997). "A Note on "A Concordance Correlation Coefficient to Evaluate Reproducibility". Biometrics (International Biometric Society) 53 (4): 1503–1507. doi:10.2307/2533516. JSTOR 2533516. 
  3. ^ Reinhold Müller & Petra Büttner (December 1994). "A critical discussion of intraclass correlation coefficients". Statistics in Medicine 13 (23–24): 2465–2476. doi:10.1002/sim.4780132310. PMID 7701147. 
  4. ^ Klaus Krippendorff (1970). "Bivariate agreement coefficients for reliability of data". In E. F. Borgatta. Sociological Methodology 2. San Francisco: Jossey-Bass. pp. 139–150. doi:10.2307/270787. 
  5. ^ Lawrence I-Kuei Lin (March 2000). "A Note on the Concordance Correlation Coefficient". Biometrics 56: 324–325. doi:10.1111/j.0006-341X.2000.00324.x. 
  6. ^ Nicholas Lange, Stephen C. Strother, J. R. Anderson, Finn Årup Nielsen, Andrew P. Holmes, Thomas Kolenda, Robert L. Savoy and Lars Kai Hansen (September 1999). "Plurality and resemblance in fMRI data analysis". NeuroImage 10 (3 Part 1): 282–303. doi:10.1006/nimg.1999.0472. PMID 10458943.