Résumés
Résumé
L’alpha de Cronbach est l’indice de consistance interne le plus répandu en sciences de l’éducation. Le but de cet article est d’évaluer la performance de six estimateurs de consistance interne à partir d’une étude de simulation. La simulation porte sur l’alpha de Cronbach, le lambda-2, le lambda-4 et le lambda-6 de Guttman, la plus grande limite inférieure et l’oméga. Quarante-cinq scénarios ont été définis par la taille de l’échantillon, le nombre d’items et la valeur des coefficients de saturation factorielle. Les résultats suggèrent que, dans le cas où l’instrument compte cinq items, l’estimateur à privilégier serait l’oméga. Dans les autres cas, ce serait la grande limite inférieure. L’alpha et le lambda-2 sont systématiquement les deux estimateurs qui sous-estiment le plus la valeur de la consistance interne et devraient être évités. Le lambda-6 serait le meilleur estimateur offert par SPSS. Dans l’ensemble, cette étude offre un rationnel empirique pour un changement de pratique dans les recherches en éducation.
Mots-clés :
- fidélité,
- consistance interne,
- mesure,
- simulation,
- alpha de Cronbach
Abstract
Cronbach’s alpha is the most common internal consistency index in educational sciences. The goal of this article is to assess the performance of six internal consistency estimators by conducting a simulation study. The simulation focuses on Cronbach’s alpha, lambda-2, Guttman’s lambda-4 and lambda-6, the greatest lower bound (GLB) and the omega. Forty-five scenarios were defined by the sample size, the number of items and the value of factorial saturation coefficients. The results suggest that if the instrument has five items, the preferred estimator would be omega. In other cases, it would be the GLB. Alpha and lambda-2 are systematically the two estimators that most underestimate the value of internal consistency and should be avoided. Lambda-6 would be the best estimator offered by SPSS. Overall, this study offers an empirical rationale for a change of practice in educational research.
Keywords:
- fidelity,
- internal consistency,
- measure,
- simulation,
- Cronbach’s alpha
Resumen
El alfa de Cronbach es el índice de consistencia interna más extendido en ciencias de la educación. El objetivo de este artículo es evaluar el rendimiento de seis estimadores de consistencia interna a partir de un estudio de simulación. La simulación trata del alfa de Cronbach, el lambda-2, el lambda-4 y el lambda-6 de Guttman, el mayor límite inferior y el omega. Se definieron cuarenta y cinco situaciones según el tamaño de la muestra, el número de ítems y el valor de los coeficientes de saturación factorial. Los resultados sugieren que, en el caso en que el instrumento cuente con cinco ítems, el estimador a privilegiar sería el omega. En los otros casos, sería el mayor límite inferior. El alfa y el lambda-2 son sistemáticamente los dos estimadores que más subestiman el valor de la consistencia interna y deberían ser evitados. El lambda-6 sería el mejor estimador ofrecido por el programa informático SPSS. De manera general, nuestro estudio ofrece una justificación empírica para un cambio de prácticas en las investigaciones en educación.
Palabras clave:
- fidelidad,
- consistencia interna,
- medida,
- simulación,
- alfa de Cronbach
Parties annexes
Bibliographie
- Bentler, P. M. (2009). Alpha, dimension-free, and model-based internal consistency reliability. Psychometrika, 74(1), 137-143.
- Callender, J. C. et Osburn, H. G. (1979). An empirical comparison of coefficient alpha, Guttman’s lambda-2, and MSPLIT maximized split-half reliability estimates. Journal of educational measurement, 16(2), 89-99.
- Cho, E. (2016). Making reliability reliable: A systematic approach to reliability coefficients. Organizational research methods, 19(4), 651-682.
- Cho, E. et Kim, S. (2015). Cronbach’s coefficient alpha: Well known but poorly understood. Organizational research methods, 18(2), 207-230.
- Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of applied psychology, 78(1), 98-104.
- Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297-334.
- Cronbach, L. J. et Shavelson, R. J. (2004). My current thoughts on coefficient alpha and successor procedures. Educational and psychological measurement, 64(3), 391-418.
- Davenport, E. C., Davison, M. L., Liou, P.-Y. et Love, Q. U. (2015). Reliability, dimensionality, and internal consistency as defined by Cronbach: Distinct albeit related concepts. Educational measurement: Issues and practice, 34(4), 1-6.
- Green, S. B. et Yang, Y. (2009). Commentary on coefficient alpha: A cautionary tale. Psychometrika, 74(1), 121-135.
- Guttman, L. (1945). A basis for analyzing test-retest reliability. Psychometrika, 10(4), 255-282.
- Henson, R. K. (2001). Understanding internal consistency reliability estimates: A conceptual primer on coefficient alpha. Measurement and evaluation in counseling and development, 34(3), 177-189.
- Hunt, T. D. et Bentler, P. M. (2015). Quantile lower bounds to reliability based on locally optimal splits. Psychometrika, 80(1), 182-195.
- Jackson, P. H. et Agunwamba, C. C. (1977). Lower bounds for the reliability of the total score on a test composed of non-homogeneous items: I: Algebraic lower bounds. Psychometrika, 42(4), 567-578.
- Kelley, K. et Pornprasertmanit, S. (2015). Confidence intervals for population reliability coefficients: Evaluation of methods, recommendations, and software for composite measures. Psychological methods, 21(1), 69-92.
- McDonald, R. P. (1978). Generalizability in factorable domains: “Domain validity and generalizability”. Educational and psychological measurement, 38(1), 75-79.
- McDonald, R. P. (1985). Factor analysis and related methods. Erlbaum.
- Osburn, H. G. (2000). Coefficient Alpha and related internal consistency reliability coefficients. Psychological methods, 5(3), 343-355.
- Revelle, W. (1979). Hierarchical cluster analysis and the internal structure of tests. Multivariate behavioral research, 14(1), 57-74.
- Revelle, W. et Zinbarg, R. E. (2009). Coefficients alpha, beta, omega, and the GLB: Comments on Sijtsma. Psychometrika, 74(1), 145-154.
- Schmitt, N. (1996). Uses and abuses of coefficient Alpha. Psychological assessment, 8(4), 350-353.
- Shapiro, A. et Ten Berge, J. M. F. (2000). The asymptotic bias of minimum trace factor analysis, with applications to the greatest lower bound to reliability. Psychometrika, 65(3), 413-425.
- Sijtsma, K. (2009a). On the use, the misuse, and the very limited usefulness of Cronbach’s alpha. Psychometrika, 74(1), 107-120.
- Sijtsma, K. (2009b). Reliability beyond theory and into practice. Psychometrika, 74(1), 169-173.
- Streiner, D. L. (2003). Starting at the beginning: An introduction to coefficient alpha and internal consistency. Journal of personality assessment, 80(1), 99-103.
- Streiner, D. L., Norman, G. R. et Cairney, J. (2015). Health measurement scales. A practical guide to their development and use (5th édition). Oxford University Press.
- Tang, W. et Cui, Y. (2012, 17 avril). A simulation study for comparing three lower bounds to reliability. Communication présentée à l’AERA Division D: Measurement and research methodology, section 1: Educational measurement, psychometrics, and assessment (1-25).
- Ten Berge, J. M. F. et Sočan, G. (2004). The greatest lower bound to the reliability of a test and the hypothesis of unidimensionality. Psychometrika, 69(4), 613-625.
- Ten Berge, J. M. F. et Zegers, F. E. (1978). A series of lower bounds to the reliability of a test. Psychometrika, 43(4), 575-579.
- Trizano-Hermosilla, I. et Alvarado, J. M. (2016). Best alternatives to Cronbach’s Alpha reliability in realistic conditions: Congeneric and asymmetrical measurements. Frontiers in psychology, 7(769), 1-8.
- Woodhouse, B. et Jackson, P. H. (1977). Lower bounds for the reliability of the total score on a test composed of non-homogeneous items: II: A search procedure to locate the greatest lower bound. Psychometrika, 42(4), 579-591.