Paulo Carvalho, Irene-Angelica Chounta, Patrick Carrington, Sang-Won Bae
Post-Doctoral Researchers, Human-Computer Interaction Institute, Carnegie Mellon University
- Friday, September 8, 2017 - 1:30pm to 2:30pm
- Newell-Simon Hall 1305 (Michael Mauldin Auditorium)
- Seminar Video
Speaker 1: Paulo Carvalho
Talk Title: Beyond reading in educational contexts: Spending more time on practice activities as a better way to learn.
Learning by doing refers to learning practices that involve completing activities as opposed to explicit learning (e.g., reading). Although the benefits of learning by doing have been described before, it is still relatively uncommon in instructional practice. We investigated how much students employ learning by doing and self-quizzing in online courses, and whether it is associated with improved learning outcomes. Spending more time completing activities had a larger impact on learning outcomes than spending more time reading, even in the case of mostly declarative content, such as in a Psychology course. Importantly, this positive effect of practice activities was seen not only for multiple-choice quizzes and exams but also for open-ended essay questions. Finally, we contrast this evidence with students’ a priori intuitions on best study strategies for their online course. Students overestimate the value of explicit learning through reading, and underestimate the value of active learning and quizzing.
Speaker 2: Irene-Angelica Chounta
Talk Title: A Computational Model of the Zone of Proximal Development: the “Grey Area” Approach.
The goal of our project is to develop an adaptive, natural-language tutoring system, driven by a student model, which can effectively carry out reflective conversations with students after they solve physics problems. Choosing the “next step” is a prominent issue in the case of dialogue-based intelligent tutors: not only should the task be appropriate with respect to the background knowledge of the student, but it should also be presented in an appropriate manner so that the student will not be overwhelmed and discouraged. Following the practice of human tutors, we propose a computational approach to model the ZPD of students who carry out learning activities using a dialogue-based intelligent tutoring system. We employ a student model to assess students’ changing knowledge as they engage in a dialogue with the system. Based on the model’s predictions, we define the concept of the “Grey Area”, a probabilistic region in which the model’s predictive accuracy is low. We argue that this region can be used to indicate whether a student is in the ZPD. Deriving ways to identify and formally describe the ZPD is an important step towards understanding the mechanisms that drive learning and development, gaining insights about learners’ needs, and providing appropriate pedagogical interventions. In this talk, I will present the concept of the “Grey Area” and how it can provide insight with respect to student’s ZPD. I will demonstrate the application of this approach on existing data and some early findings. I will present its theoretical and practical implications for ITSs and I will discuss the contribution and limitations of this work and our plans for the future.
Speaker 3: Patrick Carrington
Talk Title: Chairable Computing: Designing Mobile and Wearable Technologies for Wheelchair Users
Wheelchair users often use and carry multiple mobile computing devices. Many wheelchair users have some upper body motor impairment that can make using these devices difficult. A wheelchair user’s ability to interact with computing technology may also be physically restricted by the wheelchair’s frame, which can obstruct movement or limit reach. Designing technology for wheelchair users requires a constant negotiation between the user’s needs, social factors, technological constraints, and the context in which the technology will be used. Complex individual differences can make designing assistive solutions uniquely challenging as solutions are often designed for one individual, and do not generalize. These devices have a profound potential to alter the quality of life experiences for this population. In this talk, I will discuss my work defining the research area of Chairable Computing. I contribute to the inclusive design of technology through an ability-based approach by developing devices that expand the perception and expressive capabilities of technology for wheelchair users. I will present findings from the design and development of a wheelchair-based gesture-input device as well as exploration into the development of activity monitors for wheelchair athletes. I use these findings to discuss further design directions for chairable computing, wearable computing, and assistive technologies.
Speaker 4: Sang-Won Bae
Talk Title: How can we detect whether young adults drink alcohol solely using smartphones?
In this talk, she explores the detection of daily-life behavioral markers using mobile phone sensors and their utility in identifying drinking occasions. She utilized data from young adults with past hazardous drinking and collected mobile phone sensor data and daily questions of drinking for 28 consecutive days. She built a machine learning-based model identifying non-drinking, drinking and heavy drinking episodes. She will highlight the most important features for detecting drinking episodes and identify the amount of historical data needed for accurate detection. Her results suggest that mobile phone sensors can be used for automated, continuous monitoring of at-risk populations to detect heavy drinking episodes and support the delivery of timely interventions.
- Speaker's Bio
Paulo Carvalho is a Postdoctoral Fellow in the Human-Computer Interaction Institute at Carnegie Mellon University working with Ken Koedinger. His research investigates how the way students organize their study changes what they learn. For example, how different learning environments lead to the creation of different representations of what was studied. This knowledge can help us understand learning more broadly, create better models of learning, and to adapt learning environments to be organized in the optimal way to improve learning. He completed a MS in Experimental Psychology at Universidade do Minho in Portugal and a PhD in Psychological and Brain Sciences at Indiana University, where he studied human learning in and outside the laboratory and developed cognitive models of learning.
Since April 2016 Irene-Angelica Chounta has held a post-doctoral researcher position in the Human-Computer Interaction Institute, Carnegie Mellon University. She works with Patricia Albacete, Pamela Jordan, Sandra Katz (Learning Research and Development Center, University of Pittsburgh) and Bruce McLaren (HCII, CMU) on a joint project between CMU and the University of Pittsburgh that aims to develop a student model to support a physics tutorial dialogue system. She studied Electrical and Computer Engineering and received her Ph.D. from the University of Patras, Greece. During her Ph.D. she focused on human-computer interaction, groupware evaluation, technology-enhanced and computer-supported collaborative learning. From August 2014 until March 2016, she was a postdoctoral researcher in the Collide research group of the University of Duisburg-Essen in Germany under the supervision of Prof. H. Ulrich Hoppe. Her research activities mainly focus on learning analytics for monitoring, mirroring and guiding learning activities, group formation and group dynamics in collaborative and cooperative settings. Her main interest is to model learners’ behavior over time in order to provide feedback and scaffold learning, both in formal or informal contexts.
Patrick Carrington received his Ph.D. in Human-Centered Computing (HCC) at the University of Maryland, Baltimore County (UMBC), where he was advised by Dr. Amy Hurst. As an accessibility and human-computer interaction researcher, he studies mobile and wearable technology, builds assistive devices, and explores how technology can be used to support empowerment and independence. Patrick’s research aims to understand the needs and practices of wheelchair users to design technology systems that enhance and leverage the full potential of people with diverse abilities. He received his M.S. in HCC and B.S. in Information Systems with a certificate in Web Development from UMBC. Patrick has multiple conference and journal publications in the aforementioned topics, winning a Best Paper Honorable Mention award at CHI 2014. His work has been funded by Microsoft Research and Nokia Research. He is also a 2016 HIMSS National Capitol Area Scholar, 2015 Generation Google Scholar, and 2011-2013 National Science Foundation LSAMP Bridge to Doctorate fellow.
Sang-Won Bae is a Systems Scientist at HCI Institute at Carnegie Mellon University, mentored by Anind Dey. She is interested in predicting risky behaviors and assisting people by using mobile sensors. The goal of her work is to examine the feasibility and acceptability of collecting continuously-sensed contextual information and active patient-reported symptom reports and to use these types of information to develop algorithms to accurately predict behavior for use in health monitoring and treatment delivery such as chronic symptoms, depression, hospital readmissions and substance use and abuse such as binge drinking, marijuana use, and opioid addiction. Dr. Bae received her Ph.D. in HCI, Cognitive Science and Engineering, at Yonsei University, advised by Jin-Woo Kim who was the ACMSIG CHI Conference General Co-Chair, and her team has received several grants and awards from a number of foundations, including the National Research Foundation of Korea (NRF) and the Korea Institute of Science and Technology (KIST) in South Korea, and she has supported by the CTSI T1/T2 Pilot Awards by the National Center for Advancing Translational Sciences (NCATS) and the R21 grant from the National Institutes of Health (NIH) in the United States.
- Brad Myers