Reliability analysis is a method to assess the consistency and dependability of the responses to questions in a questionnaire that uses scales. It is used to verify if the sample has provided accurate answers to the scale questions. Reliability analysis is not typically performed on non-scale data. The most used coefficient for measuring internal consistency reliability is Cronbach’s alpha, which shows the level of consistency among the responses to the items in a scale. The formula for Cronbach’s alpha is:

The formula for Cronbach’s alpha reliability coefficient includes “k”, which represents the number of items measured, “Si^2”, which is the variance of the score for the i-th question, and “Sx^2”, which is the variance of the total test score. A reliability coefficient above 0.8 indicates that the scale is highly reliable, while a coefficient above 0.7 is acceptable. If the coefficient is above 0.6, it is necessary to modify the scale, but if it is less than 0.6, the scale requires significant changes.

References:

• Lambert, M. J., Burlingame, G. M., Umphress, V., Hansen, N. B., Vermeersch, D. A., Clouse, G. C., & Yanchar, S. C. (1996). The reliability and validity of the Outcome Questionnaire. Clinical Psychology & Psychotherapy: An International Journal of Theory and Practice, 3(4), 249-258.
• Taherdoost, H. (2016). Validity and reliability of the research instrument: How to test the validation of a questionnaire/survey in research. Electronic Physician, 8(9), 3150-3155.
• Henson, R. K., & Roberts, J. K. (2006). Use of exploratory factor analysis in published research: Common errors and some comment on improved practice. Educational and Psychological measurement, 66(3), 393-416..