ABSTRACT
This chapter considers the implications for learning the structure of a multinomial processing tree underlying a task. Multinomial processing trees have been successfully used to model processes in many tasks, including perception, memory and social cognition. Distinguishing serial and parallel processes is analogous to distinguishing ordered and unordered processes in a multinomial processing tree. The Transitive Orientation Algorithm can be used to arrange the processes in a multinomial processing tree. The set of all processes in a multinomial processing tree is partially ordered; that is, the relation of precedence between two processes is irreflexive and transitive. If the edges can be oriented in a way that the vertices are partially ordered, the orientation is called a transitive orientation, and the graph is called a comparability graph. The directed acyclic graph that results from removing all arcs that make redundant paths is called a Hasse diagram.
