Open AccessINTERPRETING STUDENT RESPONSES USING SENTIMENT ANALYSIS AND TEXT-ANALYTICS
Author(s) -
Madison Mindorff,
Brendan Lobo,
Chirag Variawa
Publication year - 2021
Publication title -
proceedings of the ... ceea conference
Language(s) - English
Resource type - Journals
ISSN - 2371-5243
DOI - 10.24908/pceea.vi0.14866
Subject(s) - sentiment analysis , learning analytics , expectancy theory , vocabulary , computer science , data science , value (mathematics) , mathematics education , online learning , analytics , psychology , artificial intelligence , world wide web , linguistics , machine learning , social psychology , philosophy
This paper discusses exploratory research which computationally examines over one and a half million words presented by first-year students as part of weekly online assignments over the Fall 2020 academic term. This work aims to explore whether computational analyses of first-year engineering student vocabulary can be employed to understand the levels of student motivation when learning engineering in an online environment. The investigation uses NVivo 12 Plus (NVivo), a data analysis software, to track the overall sentiment of weekly student discussion board responses and apply text queries to determine the number of responses that include words related to the expectancy-value theory. Applying this theory reveals trends in overall student motivation, with weeks four to six and eight to ten having an overall positive sentiment. This positive sentiment reveals higher levels of student motivation during those weeks.