Enoch O. Oladunmoye, PhD

Exploratory Factor Analysis Versus Confirmatory Factor Analysis

Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) are among the most important statistical procedures used in the development and evaluation of psychological and behavioural measurement instruments. Although the two techniques are frequently discussed together, they answer fundamentally different questions. EFA is primarily concerned with discovering or evaluating plausible latent structures when the dimensionality of an item pool is uncertain, whereas CFA evaluates a prespecified measurement model against observed data. Confusing these two purposes can lead to overfitted models, inflated claims of validity, inappropriate item deletion, and weak replication. This Research Note explains the conceptual and statistical distinction between EFA and CFA and provides a practical workflow for deciding when each procedure should be used. Particular attention is given to factor retention, item cross loadings, factor interpretation, model specification, estimator selection, model fit, cross validation, sample splitting, measurement invariance, and the relationship between exploratory findings and confirmatory evidence. The note also demonstrates how an integrated psychometric platform such as PsychtrixWeb can support the transition from exploratory measurement development to confirmatory validation, and it illustrates the discussion with a worked example, comparative tables and summary figures. The central argument is that EFA and CFA should not be regarded as competing procedures. They are complementary components of a broader measurement development process. EFA helps researchers understand the empirical structure of an item pool, while CFA subsequently tests whether a theoretically specified structure is supported by new, or appropriately independent, evidence. A robust scale development programme typically moves from theory, through exploratory investigation and refinement, to confirmatory testing, reliability and validity evaluation, and replication. Keywords: exploratory factor analysis, confirmatory factor analysis, EFA, CFA, factor analysis, scale development, psychometrics, latent variables, structural validity, measurement invariance, PsychtrixWeb