Expanding the CMR Model – A Computational Investigation of Reading Comprehension (DSI-SRP)
This DSI-SRP fellowship funded Elijah Williams to work in the laboratory of Dr. Katherine S. Aboud, Ph.D. in the Vanderbilt Brain Institute during the summer of 2023. Elijah is a rising junior with a major in Mathematics and a minor in Data Science.
In this summer research project, Elijah and his mentor investigated patterns in the immediate free recall of expository texts, with the aim of developing a model that could explain the most prominent patterns in the data, such as the temporal contiguity effect (the tendency to remember things based on the time they occurred), as well as the huge variation in how memorable ideas from each text were. To this end, they used the Context, Maintenance, and Retrieval (CMR) model of memory retrieval (Polyn, Norman, Kahana, 2009), alongside measures of text properties such as concreteness and arousal, to predict how and what ideas people would remember from each text. When fitted to their behavioral data, they found the neural network and regression model could account for several aspects of the shape of the serial position and lag conditional-response probability (lag-CRP) curves. In the future, they plan on combining these two models into a single model of text comprehension and retrieval, as well as exploring the role that schema and passage structure play when learning new information from a text. Elijah would like to thank the DSI, whose workshops and sessions involving NLP, Jupyter notebooks, and Python in general, were extremely handy when it came to first exploring and analyzing the free recall data.
In addition to receiving support through a DSI-SRP fellowship, this project was supported and facilitated by the DSI Data Science Team through their regular summer workshops and demo sessions.