Language Perception Laboratory

Software

TRACE
The .zip file contains primarily the files that were in the orginial release provided by Jeff Elman, plus a few extras, such as some C programs that I use to filter and condense the output. Use them and this version of TRACE at your own risk.

Although the functionality of the original instantiation has not been modified, documented and undocumented changes to the program have been made when porting it from one platform to the next. It has run successfully under Solaris and is currently in use under Linux.

All this said, if all you want to do is run simulations, download the new Java version of TRACE from here. It is quite user friendly. You will be up and running in no time. Jim Magnuson and his team have provided a great service to the psycholinguistic community.
TRACE

Adaptive Resonance Theory (ART)
This file contains matlab code for the version of ARTphone used in Pitt, Myung, and Altieri (2007), along with a more detailed description of the model, including equations, than is presented in the paper.
ARTphone.zip

Parameter Space Partition
PSP is a search algorithm that can be used to discover the number of qualitatively different data patterns a model can generate in an experimental design, which can serve as an indexo of its flexibility. It carves up the parameter space of a model into regions (i.e., ranges of parameter settings) that correspond to different data patterns. A demonstration of the algorithm along with Matlab source code can be found here.

Design Optimization
Advances in understanding behavior often depend on well designed experiments. The choices of design variables, and especially their levels (e.g., how many conditions and how are they spaced), can significantly affect the results, and thus the informativeness of the experiment. When comparing mathematical models, knowledge of model behavior can be used to identify a design that will yield the highest probability of discriminating among the models. Design Optimization is a formal method of achieving this goal. The C++ code that was used in its application to retention and categorization models is available here.