Student Co-presenter Names

Daniel Carstensen

Download

Download Full Text (3.3 MB)

Student Class

2024

Student Affiliation

Senior Honors Thesis

Author ORCID Identifier

https://orcid.org/0000-0001-7613-4732

First Advisor

Jeremy Manning

First Advisor Department

Department of Psychological and Brain Sciences

Second Advisor

Peter Mucha

Second Advisor Department

Department of Mathematics

Description

This study investigates the potential of a computational approach to provide moment-by-moment insights into a student’s comprehension of lecture material through analysis of their neurophysiological responses during the lecture. In doing so, we present a solution to two difficult problems. How do we quantify the conceptual content of a lecture video? And how can we use EEG recordings to compute a knowledge estimate of this conceptual content? First, we used topic modeling to generate moment-by-moment estimates of the conceptual content of a lecture. Then we used EEG recordings collected during the lecture to compute an ISFC-derived knowledge estimate of this conceptual content. We found that particularly gamma band activity may contain a signal indicative of knowledge acquisition.

Publication Date

5-22-2024

Keywords

computational neuroscience, education, learning, knowledge, concepts, natural language processing, eeg

Disciplines

Cognitive Neuroscience | Computational Neuroscience | Statistical Models

Translating Neurophysiological Recordings Into Dynamic Estimates of Conceptual Knowledge And Learning

Restricted

Available to Dartmouth community via local IP address.

Share

COinS