Asymptotically Distribution Free (Adf) Interval Estimation of Coefficient Alpha

Posted: 20 Sep 2007

See all articles by Donna L. Coffman

Donna L. Coffman

affiliation not provided to SSRN

Alberto Maydeu Olivares

Fundación Instituto de Empresa, S.L.

Date Written: May 12, 2006

Abstract

Asymptotic distribution free (ADF) interval estimators for coefficient alpha were introduced in the context of an application by Yuan, Guarnaccia, and Hayslip (2003). Here, simulation studies were performed to investigate the behavior of ADF vs. normal theory (NT) interval estimators of coefficient alpha for tests composed of ordered categorical items under varied conditions of sample size, item skewness and kurtosis, number of items, and average inter-item correlation. NT intervals were found to be inaccurate when item skewness > 1 or kurtosis > 4. But for sample sizes over 100 observations, ADF intervals provide an accurate perspective of the population coefficient alpha of the test regardless of item skewness and kurtosis. A formula for computing ADF confidence intervals for coefficient alpha for tests of any size is provided, along with its implementation as a SAS macro.

Keywords: coefficient omega, reliability, Likert-type ítems.

JEL Classification: C00

Suggested Citation

Coffman, Donna L. and Maydeu Olivares, Alberto, Asymptotically Distribution Free (Adf) Interval Estimation of Coefficient Alpha (May 12, 2006). Instituto de Empresa Business School Working Paper No. WP06-24, Available at SSRN: https://ssrn.com/abstract=1015835

Donna L. Coffman

affiliation not provided to SSRN ( email )

No Address Available

Alberto Maydeu Olivares (Contact Author)

Fundación Instituto de Empresa, S.L. ( email )

Mª Molina, 11,13,15
Madrid, Madrid 28006
Spain
915 689 732 (Phone)

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
577
PlumX Metrics