Artificial grammar learning
We consider Artificial Grammar Learning (AGL), which is a versatile methodological tool for the study of learning. AGL is fairly unique amongst learning paradigms, in that it allows an instantiation of a wide variety of theories of learning, including rules, similarity, and associative learning theories. Also, performance in AGL tasks typically reflects both implicit and explicit learning processes. We review these putative influences on AGL performance and how they relate to general cognitive theory. This flexibility of the AGL paradigm comes at a price, in that sophisticated modeling and analytical methods are required in order to make precise hypotheses about the psychological basis of AGL performance in particular cases. We review methodological issues and briefly assess a range of analytical tools in AGL.