Week 3 - Reliability Flashcards
Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78, 98-104.
The paper by Cortina (1993) titled “What is coefficient alpha? An examination of theory and applications” in the Journal of Applied Psychology explores the concept of coefficient alpha, a crucial statistic in test construction and research. This paper addresses the widespread use and confusion regarding the true meaning and interpretation of coefficient alpha. Cortina clarifies its theoretical aspects, integrating various statements made about alpha, and takes a practical approach by examining how alpha changes with the manipulation of item numbers and inter-item correlations. This study is pivotal in understanding the application and interpretation of coefficient alpha in a wide range of fields beyond psychology, including sociology, medicine, economics, and more (Cortina, 1993).
McNeish, D. (2018). Thanks coefficient alpha, we’ll take it from here. Psychological Methods, 23(3), 412.
The paper by McNeish (2018), titled “Thanks coefficient alpha, we’ll take it from here,” in Psychological Methods, critiques the widespread use of Cronbach’s alpha as a measure of internal consistency reliability in psychological research. McNeish argues that despite its popularity, Cronbach’s alpha is flawed due to unrealistic assumptions, often leading to underestimations of reliability. The paper provides evidence that methodological critiques of alpha have not significantly changed its use in empirical studies. McNeish conceptually outlines the drawbacks of Cronbach’s alpha, avoiding technical or mathematical arguments to make it accessible to a broader audience. The paper suggests alternative measures that make less rigid assumptions and offer more accurate reliability estimates than alpha. These alternatives include omega total, Revelle’s omega total, the greatest lower bound, and Coefficient H. The paper concludes with empirical examples demonstrating the advantages of these alternatives and includes a software appendix for implementation (McNeish, 2018).
Davenport, E. C., Davison, M. L., Liou, P. Y., & Love, Q. U. (2015). Reliability, dimensionality, and internal consistency as defined by Cronbach: Distinct albeit related concepts. Educational Measurement: Issues and Practice, 34(4), 4-9.
The paper by Davenport, Davison, Liou, and Love (2015), titled “Reliability, Dimensionality, and Internal Consistency as Defined by Cronbach: Distinct Albeit Related Concepts” in Educational Measurement: Issues and Practice, clarifies the concepts of reliability, dimensionality, and internal consistency, distinguishing them as separate yet interrelated. The authors start by critiquing the definition of reliability and then explore the mathematical properties of Cronbach’s alpha. They discuss internal consistency and dimensionality as defined by Cronbach and present functional relationships among reliability, internal consistency, and dimensionality. The paper demonstrates the utility of these concepts and suggests that reliability, internal consistency, and dimensionality should each be quantified with separate indices, acknowledging their interrelatedness. The authors argue that high levels of unidimensionality and internal consistency are not necessary for reliability as measured by alpha or for the interpretability of test scores (Davenport, Davison, Liou, & Love, 2015).
