Original Research

Coefficient alpha: Unnecessarily ambiguous; unduly ubiquitous

G K Huysamen
SA Journal of Industrial Psychology | Vol 32, No 4 | a242 | DOI: https://doi.org/10.4102/sajip.v32i4.242 | © 2006 G K Huysamen | This work is licensed under CC Attribution 4.0
Submitted: 29 October 2006 | Published: 29 October 2006

About the author(s)

G K Huysamen, University of Pretoria and University of the Free State, South Africa

Full Text:

PDF (131KB)

Abstract

By means of simple numerical examples it is demonstrated how the value of coefficient alpha is determined by the size and sign of the item intercorrelations, the dimensionality of the items, and the number of items. Relatively high alphas may be obtained for multidimensional item data (as obtained under a violation of essential tau-equivalence). Dimensionality analyses should therefore precede internal-consistency analyses if unidimensional tests are required. Such analyses also may alert one to the possibility of negative alpha values. Alternative coefficients are mentioned for situations when essential tau-equivalence is in doubt.

Keywords

Coefficient alpha; Internal consistency

Metrics

Total abstract views: 6930
Total article views: 8153

 

Crossref Citations

1. Development and preliminary validation of the work-unit performance questionnaire
Jonathan Seland, Carl C. Theron
South African Journal of Economic and management Sciences  vol: 24  issue: 1  year: 2021  
doi: 10.4102/sajems.v24i1.3926