Contributions to the Discussion on the Assessment of the Reliability of a Measurement Instrument

Abstract

Theoretical framework: The reliability of an instrument refers to the consistency of the results obtained during its administration. Nowadays, reliability assessment is almost imperative in the presentation of empirical data. Depending on the purpose of the presentation and the concept measured, the study of reliability can include multiple procedures. Objectives: To map the main techniques for measuring reliability and their algorithms. Main topics of analysis: The test-retest and equivalent forms are used to measure temporal stability. Cronbach’s alpha, Kuder-Richardson and Spearman-Brown split-half models are used for measuring internal consistency. Conclusion: Since the study of reliability of a scale is directly related to what we intend to measure or compare, it is impossible to postulate standards for the use of the above-mentioned estimators. For that reason, we propose a careful consideration in choosing the most suitable estimator for the study to be developed.