Density of representations yields age-related dissociations (DRYAD) model

Benjamin's (2010) DRYAD model

• Aaron Benjamin, University of Illinois at Urbana-Champaign

Citation:

• Benjamin, A. S. (2010). Representational explanations of “process” dissociations in recognition: The DRYAD theory of aging and memory judgments. Psychological Review, 117(4), 1055.

DEPENDENCIES

While the goal is to provide an implementation of DRYAD that highly accessible and relatively free of dependencies, there are a few that are required to run this simulation.

• REQUIRED
  • Python 3.5+
  • NumPy (available through Anaconda suite or through Python; pip install numpy or pip3 install numpy
• OPTIONAL
  • Progress Bar (conda install progressbar2, pip install progressbar2, or pip3 install progressbar2)

MAJOR UPDATES

(092818)[TC]

• Code was continuing to under-estimate hit rate, but this error has been fixed. The code was only handling up to 6 context nodes - it can now handle much more than that.
• I am pretty confident that this is the final round of major code changes. Will be running Markov Chain analyses to compare to MATLAB output.

(091418)[TC]

• Finished annotating the helper functions in dryad_modules.py. They will continue to be revised in order to better reflect what they are doing.
• Pushed changes to the response module in dryad_modules.py. It was under-estimating the hit rate due to a coding error in which it was evaluating hits to the second context - not the first.

(090718)[TC]

• Modified the program to perform a grid search (i.e. loop through parameters). Also added a progress bar that shows how much time has elapsed as well as the approximate time left in the simulation. The progress bar can be suppressed by setting see_progress_bar to 0.
• The progress bar requires installation of some outside packages. If you running the Anaconda/Spyder suite (which is recommended), pull up the Anaconda prompt and type in conda install progressbar2. If you are running Python, pull up a Python prompt window and type in either pip install progressbar2 or pip3 install progressbar1, depending on which version of Python you are running.

(090618)[TC]

• Finsihed what I believe is a full representation of the DRYAD model. This includes a main file by which parameters can be "tuned", and a second file that holds all helper functions (represented by different MATLAB files in Benjamin's original code).
• Still need to annotate the helper functions in order for coders to line up the original code with the new Python code.