Appropriate Sample Size for Comparing Alpha Reliabilities

Abstract

In many measurement settings, it might be appropriate to test the significance of the difference between the internal consistency reliability coefficients of two competing measurement procedures. Determination of an adequate sample size is a crucial factor in the design of such studies. A graphical method was developed for determining sample size based on the power of Feldt's (1969) test of the difference between two values of Cronbach's alpha (a) coefficient. Implementation of this approach presumes that the researcher can specify a desired power for the statistical test against a particular alternative to the null hypothesis.