In terms of our self-esteem measure, examining the instrument through the lens of Classical Test Theory (CTT) versus Item Response Theory (IRT) would yield different insights regarding its reliability and validity. From a CTT perspective, the focus is on total test scores and the assumption that each observed score is composed of a true score and random error (Crocker & Algina, 2006). In this framework, reliability is often evaluated using internal consistency measures such as Cronbach’s alpha, and item statistics like item-total correlations are used to assess how well each item contributes to the overall scale. However, these statistics are sample-dependent, meaning the properties of the items may change with different populations (DeVellis, 2017).

In contrast, IRT takes a probabilistic approach and models the relationship between a person’s latent trait (e.g., self-esteem) and the probability of endorsing specific item responses (Embretson & Reise, 2000). Unlike CTT, IRT provides item-level information, such as difficulty and discrimination parameters, and can evaluate how well each item functions across varying levels of self-esteem. This theory also allows for the development of more adaptive and precise assessments, as items can be tailored to individual respondents. Moreover, IRT parameters are invariant across populations, assuming model fit, making it a more robust choice for generalizing findings and refining test items (Hambleton, Swaminathan, & Rogers, 1991).

References:

• Crocker, L., & Algina, J. (2006). Introduction to classical and modern test theory. Cengage Learning.
• DeVellis, R. F. (2017). Scale development: Theory and applications (4th ed.). Sage Publications.
• Embretson, S. E., & Reise, S. P. (2000). Item response theory for psychologists. Lawrence Erlbaum Associates.
• Hambleton, R. K., Swaminathan, H., & Rogers, H. J. (1991). Fundamentals of item response theory. Sage Publications.