Correlational research cannot establish a causal relationship between variables, and is limited by the potential influence of confounding variables. Furthermore, it cannot establish the direction of the relationship or the mechanism by which the relationship occurs.