ScholarLens Back to search
2016•The Journal of Nervous and Mental DiseaseRequires access

Assessing Binary Diagnoses of Bio-behavioral Disorders

Domenic V. Cicchetti, Ami Klin, Fred R. Volkmar

Open publisher page 17 citations

Abstract

Cohen's (Educ Psychol Meth. 1960;23:37-40) kappa statistic has been criticized recently for providing low chance-corrected coefficients (<0.40) despite high levels of examiner agreement (e.g., ≥85%). The AC1 statistic (Series. 2002;2:1-9; BMC Med Res Methodol. 2013;13:61), the alpha statistic (J Clin Epidemiol. 1996;49:775-782), the B statistic (BMC Med Res Methodol. 2013;13:97), and the delta statistic (Commun Statist Theory Meth. 2008;37:760-772) were developed to "correct" this "problem." Each of these statistics produces inaccurate results because they either derive from levels of chance agreement that are incorrect (Gwet's AC1 statistic) or fail to correct for chance levels of agreement (the alpha, B, and delta approaches). Implications for diagnoses of psychiatric disorders are critically discussed. The importance of kappa is demonstrated in a number of clinical scenarios.

About this research paper

What this paper is about

Cohen's (Educ Psychol Meth. 1960;23:37-40) kappa statistic has been criticized recently for providing low chance-corrected coefficients (<0.40) despite high levels of examiner agreement (e.g., ≥85%). The AC1 statistic (Series. 2002;2:1-9; BMC Med Res Methodol. 2013;13:61), the alpha statistic (J Clin Epidemiol. 1996;49:775-782), the B statistic (BMC Med Res Methodol. 2013;13:97), and the delta statistic (Commun Statist Theory Meth. 2008;37:760-772) were developed to "correct" this "problem." Each of these statistics produces inaccurate results because they either derive from levels of chance agreement that are incorrect (Gwet's AC1 statistic) or fail to correct for chance levels of agreement (the alpha, B, and delta approaches). Implications for diagnoses of psychiatric disorders are critically discussed. The importance of kappa is demonstrated in a number of clinical scenarios.

Why it matters

OpenAlex reports 17 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Cohen's (Educ Psychol Meth. 1960;23:37-40) kappa statistic has been criticized recently for providing low chance-corrected coefficients (<0.40) despite high levels of examiner agreement (e.g., ≥85%). The AC1 statistic (Series. 2002;2:1-9; BMC Med Res Methodol. 2013;13:61), the alpha statistic (J Clin Epidemiol. 1996;49:775-782), the B statistic (BMC Med Res Methodol. 2013;13:97), and the delta statistic (Commun Statist Theory Meth. 2008;37:760-772) were developed to "correct" this "problem." Each of these statistics produces inaccurate results because they either derive from levels of chance agreement that are incorrect (Gwet's AC1 statistic) or fail to correct for chance levels of agreement (the alpha, B, and delta approaches). Implications for diagnoses of psychiatric disorders are critically discussed. The importance of kappa is demonstrated in a number of clinical scenarios.

Key concepts: Statistic, Cohen's kappa, Statistics, Medical diagnosis, Kappa, Scan statistic, Mathematics, Medicine

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
Assessing Binary Diagnoses of Bio-behavioral Disorders — Research Paper | ScholarLens