How does CFA differ from EFA?

CFA (confirmatory factor analysis) and EFA (exploratory factor analysis) are both techniques for understanding the structure of data and for analyzing complex relationships among multiple variables. CFA is mainly used for model testing and validation, allowing the researcher to confirm or reject pre-specified hypotheses and interpret a model's ability to explain the data at hand. This method generally relies on utilizing and testing a specific model. EFA is mainly used for model creation and exploration, allowing the investigator to discover patterns in the data and identify underlying latent variables and their relationships. This method generally utilizes a more open-ended process and relies more heavily on statistical criteria. In sum, CFA is used to test a preconceived model, while EFA is used to explore the data and identify patterns.