Limits...
cocor: a comprehensive solution for the statistical comparison of correlations.

Diedenhofen B, Musch J - PLoS ONE (2015)

Bottom Line: The package also includes an implementation of Zou's confidence interval for all of these comparisons.The platform independent cocor package enhances the R statistical computing environment and is available for scripting.Two different graphical user interfaces-a plugin for RKWard and a web interface-make cocor a convenient and user-friendly tool.

View Article: PubMed Central - PubMed

Affiliation: Department of Experimental Psychology, University of Duesseldorf, Duesseldorf, Germany.

ABSTRACT
A valid comparison of the magnitude of two correlations requires researchers to directly contrast the correlations using an appropriate statistical test. In many popular statistics packages, however, tests for the significance of the difference between correlations are missing. To close this gap, we introduce cocor, a free software package for the R programming language. The cocor package covers a broad range of tests including the comparisons of independent and dependent correlations with either overlapping or nonoverlapping variables. The package also includes an implementation of Zou's confidence interval for all of these comparisons. The platform independent cocor package enhances the R statistical computing environment and is available for scripting. Two different graphical user interfaces-a plugin for RKWard and a web interface-make cocor a convenient and user-friendly tool.

No MeSH data available.


A flowchart of how to use the four main functions of cocor, displaying all available tests.For each case, an example of the formula passed as an argument to the cocor() function and the required correlation coefficients for the functions cocor.indep.groups(), cocor.dep.groups.overlap(), and cocor.dep.groups.nonoverlap() are given. The test label before the colon may be passed as a function argument to calculate specific tests only.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4383486&req=5

pone.0121945.g001: A flowchart of how to use the four main functions of cocor, displaying all available tests.For each case, an example of the formula passed as an argument to the cocor() function and the required correlation coefficients for the functions cocor.indep.groups(), cocor.dep.groups.overlap(), and cocor.dep.groups.nonoverlap() are given. The test label before the colon may be passed as a function argument to calculate specific tests only.

Mentions: With cocor (version 1.1-0), we provide a comprehensive solution to compare two correlations based on either dependent or independent groups. The cocor package enhances the R programming environment [33], which is freely available for Windows, Mac, and Linux systems and can be downloaded from CRAN (http://cran.r-project.org/package=cocor). All that is needed to install the cocor package is to type install.packages(“cocor”) in the R console, and the functionality of the package is made available by typing library(“cocor”). The function cocor() calculates and compares correlations from raw data. The underlying variables are specified via a formula interface (see Fig. 1). If raw data are not available, cocor offers three functions to compare correlation coefficients that have already been determined. The function cocor.indep.groups() compares two independent correlations, whereas the functions cocor.dep.groups.overlap() and cocor.dep.groups.nonoverlap() compare two dependent overlapping or nonoverlapping correlations, respectively. Internally, cocor() passes the calculated correlations coefficients to one of these three functions. All functions allow to specify the argument .value to test whether the difference between the correlations exceeds a given threshold using the confidence intervals by Zou [16]. The results are either returned as an S4 object of class cocor whose input and result parameters can be obtained using the get.cocor.input() and get.cocor.results() functions, respectively. Optionally, results may also be returned as a list of class htest. By default, all tests available are calculated. Specific tests can be selected by passing a test label to the function using the test argument. The flowchart in Fig. 1 shows how to access the available tests and lists them with their individual test label (e.g., zou2007). The formulae of all implemented tests are detailed in S1 Appendix.


cocor: a comprehensive solution for the statistical comparison of correlations.

Diedenhofen B, Musch J - PLoS ONE (2015)

A flowchart of how to use the four main functions of cocor, displaying all available tests.For each case, an example of the formula passed as an argument to the cocor() function and the required correlation coefficients for the functions cocor.indep.groups(), cocor.dep.groups.overlap(), and cocor.dep.groups.nonoverlap() are given. The test label before the colon may be passed as a function argument to calculate specific tests only.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4383486&req=5

pone.0121945.g001: A flowchart of how to use the four main functions of cocor, displaying all available tests.For each case, an example of the formula passed as an argument to the cocor() function and the required correlation coefficients for the functions cocor.indep.groups(), cocor.dep.groups.overlap(), and cocor.dep.groups.nonoverlap() are given. The test label before the colon may be passed as a function argument to calculate specific tests only.
Mentions: With cocor (version 1.1-0), we provide a comprehensive solution to compare two correlations based on either dependent or independent groups. The cocor package enhances the R programming environment [33], which is freely available for Windows, Mac, and Linux systems and can be downloaded from CRAN (http://cran.r-project.org/package=cocor). All that is needed to install the cocor package is to type install.packages(“cocor”) in the R console, and the functionality of the package is made available by typing library(“cocor”). The function cocor() calculates and compares correlations from raw data. The underlying variables are specified via a formula interface (see Fig. 1). If raw data are not available, cocor offers three functions to compare correlation coefficients that have already been determined. The function cocor.indep.groups() compares two independent correlations, whereas the functions cocor.dep.groups.overlap() and cocor.dep.groups.nonoverlap() compare two dependent overlapping or nonoverlapping correlations, respectively. Internally, cocor() passes the calculated correlations coefficients to one of these three functions. All functions allow to specify the argument .value to test whether the difference between the correlations exceeds a given threshold using the confidence intervals by Zou [16]. The results are either returned as an S4 object of class cocor whose input and result parameters can be obtained using the get.cocor.input() and get.cocor.results() functions, respectively. Optionally, results may also be returned as a list of class htest. By default, all tests available are calculated. Specific tests can be selected by passing a test label to the function using the test argument. The flowchart in Fig. 1 shows how to access the available tests and lists them with their individual test label (e.g., zou2007). The formulae of all implemented tests are detailed in S1 Appendix.

Bottom Line: The package also includes an implementation of Zou's confidence interval for all of these comparisons.The platform independent cocor package enhances the R statistical computing environment and is available for scripting.Two different graphical user interfaces-a plugin for RKWard and a web interface-make cocor a convenient and user-friendly tool.

View Article: PubMed Central - PubMed

Affiliation: Department of Experimental Psychology, University of Duesseldorf, Duesseldorf, Germany.

ABSTRACT
A valid comparison of the magnitude of two correlations requires researchers to directly contrast the correlations using an appropriate statistical test. In many popular statistics packages, however, tests for the significance of the difference between correlations are missing. To close this gap, we introduce cocor, a free software package for the R programming language. The cocor package covers a broad range of tests including the comparisons of independent and dependent correlations with either overlapping or nonoverlapping variables. The package also includes an implementation of Zou's confidence interval for all of these comparisons. The platform independent cocor package enhances the R statistical computing environment and is available for scripting. Two different graphical user interfaces-a plugin for RKWard and a web interface-make cocor a convenient and user-friendly tool.

No MeSH data available.