An Overview of Coefficient Alpha and a Reliability Matrix for Estimating Adequacy of Internal Consistency Coefficients with Psychological Research Measures

Abstract

The present article addresses issues in reliability assessment that are often neglected in psychological research such as acceptable levels of internal consistency for research purposes, factors affecting the magnitude of coefficient alpha (α), and considerations for interpreting α within the research context. A new reliability matrix anchored in classical test theory is introduced to help researchers judge adequacy of internal consistency coefficients with research measures. Guidelines and cautions in applying the matrix are provided.