College of Computing and Digital Media Dissertations

Date of Award

Fall 12-5-2023

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

School

School of Computing

First Advisor

Peter Hastings, PhD

Second Advisor

Ryan Baker, PhD

Third Advisor

Clark Elliott, PhD

Fourth Advisor

Craig Miller, PhD

Abstract

Learning a foreign language entails cognitive and emotional obstacles. It involves complicated mental processes that affect learning and emotions. Positive emotions such as motivation, encouragement, and satisfaction increase learning achievement, while negative emotions like anxiety, frustration, and confusion may reduce performance. Foreign Language Anxiety (FLA) is a specific type of anxiety accompanying learning a foreign language. It is considered a main impediment that hinders learning, reduces achievements, and diminishes interest in learning.

Detecting FLA is the first step toward reducing and eventually overcoming it. Previously, researchers have been detecting FLA using physical measurements and self-reports. Using physical measures is direct and less regulated by the learner, but it is uncomfortable and requires the learner to be in the lab. Employing self-reports is scalable because it is easy to administer in the lab and online. However, it interrupts the learning flow, and people sometimes respond inaccurately. Using sensor-free human behavioral metrics is a scalable and practical measurement because it is feasible online or in class with minimum adjustments.

To overcome FLA, researchers have studied the use of robots, games, or intelligent tutoring systems (ITS). Within these technologies, they applied soothing music, difficulty reduction, or storytelling. These methods lessened FLA but had limitations such as distracting the learner, not improving performance, and producing cognitive overload. Using an animated agent that provides motivational supportive feedback could reduce FLA and increase learning.

It is necessary to measure FLA effectively with minimal interruption and then successfully reduce it. In the context of an e-learning system, I investigated ways to detect FLA using sensor-free human behavioral metrics. This scalable and practical method allows us to recognize FLA without being obtrusive. To reduce FLA, I studied applying emotionally adaptive feedback that offers motivational supportive feedback by an animated agent.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.