Abstract: We demonstrate how the hierarchical models of Van der Linden (and collaborators), the diffusion type
process models of Tuerlinckx (and collaborators) and the scoring rule based models of Maris (and
collaborators) can be cast as network models. In the framework of network psychometrics, accuracy and
time are conceived of as separate nodes, and statistical dependencies are represented by edges
between them. Even though the three classes of models for response time and accuracy are
conceptually distinct, they share a common network representation in which the differences between
the models is coded in the values (function of response time) the response time nodes can take. The
network perspective offers a unified framework, which can easily be extended to capture more intricate
statistical dependencies (beyond one dimensional constructs) and more diverse sources of information
(beyond accuracy and time).

Keywords: diffusion type process models, hierarchical models, network psychometrics, scoring rule based models