Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Improving Mobile MOOC Learning via Implicit Physiological Signal Sensing

Xiao, Xiang (2017) Improving Mobile MOOC Learning via Implicit Physiological Signal Sensing. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF (Xiang Xiao's ETD)
Primary Text

Download (2MB) | Preview

Abstract

Massive Open Online Courses (MOOCs) are becoming a promising solution for delivering high- quality education on a large scale at low cost in recent years. Despite the great potential, today’s MOOCs also suffer from challenges such as low student engagement, lack of personalization, and most importantly, lack of direct, immediate feedback channels from students to instructors. This dissertation explores the use of physiological signals implicitly collected via a "sensorless" approach as a rich feedback channel to understand, model, and improve learning in mobile MOOC contexts.

I first demonstrate AttentiveLearner, a mobile MOOC system which captures learners' physiological signals implicitly during learning on unmodified mobile phones. AttentiveLearner uses on-lens finger gestures for video control and monitors learners’ photoplethysmography (PPG) signals based on the fingertip transparency change captured by the back camera. Through series of usability studies and follow-up analyses, I show that the tangible video control interface of AttentiveLearner is intuitive to use and easy to operate, and the PPG signals implicitly captured by AttentiveLearner can be used to infer both learners’ cognitive states (boredom and confusion levels) and divided attention (multitasking and external auditory distractions).

Building on top of AttentiveLearner, I design, implement, and evaluate a novel intervention technology, Context and Cognitive State triggered Feed-Forward (C2F2), which infers and responds to learners’ boredom and disengagement events in real time via a combination of PPG-based cognitive state inference and learning topic importance monitoring. C2F2 proactively reminds a student of important upcoming content (feed-forward interventions) when disengagement is detected. A 48-participant user study shows that C2F2 on average improves learning gains by 20.2% compared with a non-interactive baseline system and is especially effective for bottom performers (improving their learning gains by 41.6%).

Finally, to gain a holistic understanding of the dynamics of MOOC learning, I investigate the temporal dynamics of affective states of MOOC learners in a 22 participant study. Through both a quantitative analysis of the temporal transitions of affective states and a qualitative analysis of subjective feedback, I investigate differences between mobile MOOC learning and complex learning activities in terms of affect dynamics, and discuss pedagogical implications in detail.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Xiao, Xiangxix22@pitt.eduxix22
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairWang, Jingtaojingtaow@pitt.edujingtaow
Committee MemberBrusilovsky, Peterpeterb@pitt.edupeterb
Committee MemberHauskrecht, Milosmilos@pitt.edumilos
Committee MemberLee, Adamadamlee@pitt.eduadamlee
Date: 31 March 2017
Date Type: Publication
Defense Date: 26 September 2016
Approval Date: 31 March 2017
Submission Date: 7 March 2017
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 187
Institution: University of Pittsburgh
Schools and Programs: Dietrich School of Arts and Sciences > Computer Science
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Heart Rate; Intelligent Tutoring Systems; Physiological Signals; Affective Computing; Mobile Interfaces
Date Deposited: 31 Mar 2017 16:49
Last Modified: 02 Jul 2017 21:01
URI: http://d-scholarship.pitt.edu/id/eprint/30942

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item