profile photo
Qiaosi (Chelsea) Wang

Ph.D. Student
Human-Centered Computing
Georgia Tech
Email: qswang at gatech dot edu

I am a second-year PhD student in Human-Centered Computing at Georgia Tech specializing in Cognitive Science and Human-Computer Interaction. I'm working with Dr. Ashok Goel at the Design&Intelligence Lab (DILab). My research focuses on human-AI interaction, specifically I explore the understanding, use, and impact of AI in educational context. Recently, I am exploring the design and evaluation of an virtual AI agent to help distance learners to build online communities, with the goal of improving online learning engagement and encouraging online community of practice. My previous research focused on health informatics and patient-provider collaboration.

I received my bachelor of science degree from University of Washington, Seattle in informatics and psychology. I worked closely with the Department of Human-Centered Design and Engineering, the Information School, the DUB group, and the psychology department on numerous research projects during my undergraduate study. At UW, I worked in personal health informatics, patient-provider collaboration, user value scale, computing education, disaster information management, and health and risky behaviors. You can learn more about my past research projects below.

In my free time, I like to frequent gym, plant nurseries, cat cafes, and ceramic-painting shops.

PUBLICATION

Conference Proceedings
Design in the HCI Classroom: Setting a Research Agenda. [PDF]
Lauren Wilcox, Betsy Disalvo, Dick Henneman, Qiaosi Wang. In Proceedings of the ACM Conference on Designing Interactive Systems (DIS ‘19). ACM, San Diego, CA, USA. https://doi.org/10.1145/3322276.3322381. (Acceptance rate: 25%)
Best Paper Award (Top 1% of papers)
Journal Articles
Identifying and Planning for Change: Patient-provider Collaboration Using Lightweight Food Diaries in Healthy Eating and Irritable Bowel Syndrome. [PDF]
Chia-Fang Chung, Qiaosi Wang, Jessica Schroeder, Allison Cole, Jasmine Zia, James Fogarty, Sean A. Munson. PACM Interact. Mob. Wearable Ubiquitous Technol. 3, 1, Article 7 (March 2019), 23 pages. https://doi.org/10.1145/3314394
Posters and Presentations
21st annual Undergraduate Research Symposium, University of Washington
Novel Photo-Based Food Diaries to Support Patient-Provider Collaboration. Oral presentation. May, 2018.
2018 HCDE Research Showcase, University of Washington
Foodprint: Supporting Better Food-related Data Generation & Sharing. Poster presentation. February, 2018.
19th annual Undergraduate Research Symposium, University of Washington
Tonic Immobility and Maladaptive Cognitions as Predictors of Sexual Revictimization among College Women. Poster presentation. May, 2016.
Others
Mobile Apps for Generating and Sharing Food-Related Data. Collaborative Healthcare Using Patient-Generated Data.
Munson S, Chung C, Kientz J, Fogarty J, Zia J, Cole A, Schroeder J, Karkar R, Wang Q, Vilardaga R.Reos Partners, Robert Wood Johnson Foundation. December 2017. Retrieved from: https://reospartners.com/projects/collaborative-healthcare-using-patient-generated-data/

RESEARCH

Active Projects


online communities
Designing and Evaluating AI agent for Community Building in Online Education
Collaborators: Ida Camacho, David Joyner, Ashok Goel
Georgia Tech School of Interactive Computing

We are designing and developing an AI agent to help online learners build communities and social connections. We are interested in whether the community-building AI agent can help improve students’ feeling of social presence. I conducted literature review on community building in online learning, situated learrning, social cognitive theories; Designing experiments to evaluate community-building agent; Conducting interviews to understand online learners’ current community-building practices.

human-AI interaction through conversation
Evaluating Longitudinal Changes in Students' Mental Model and Interaction with Virtual Teaching Assistants
Collaborators: Marissa Gonzales, David Joyner, Ashok Goel
Georgia Tech School of Interactive Computing

We are exploring the longitudinal changes in students’ perceptions and interactions about a virtual teaching assistant (VTA) operating on Piazza forum answering students‘ questions about the class. I conducted literature review on human-AI interaction and technology perceptions; Composed IRB protocol; Designed bi-weekly surveys to measure students’ perception about the agent; Conducted statistical analysis to infer changes of students’ perception about the agent; Using mixed-methods approach to analyze verbal interactions between students and the agent to infer longitudinal changes in human-AI online interactions.

Past Projects


kids making phone call
Leveraging Behavioral and Physiological Feedback in the Design of Affect-Sensitive Distance Learning
Collaborators: Hong Li, Lauren Wilcox, Thomas Ploetz, Betsy DiSalvo
Georgia Tech School of Interactive Computing

Worked on exploring the feasibility of leveraging wearable technology and other types of sensing data to gather more context about online learners. Specifically, I am interested in the design of affect-sensi- tive wearables and the ethical and privacy issues with student affect data use in educational settings.- Composed IRB protocols and led study design; Designed and conducted in-class observational study to provide ground truth for affect data analysis; Conducted 11 semi-structured interviews with students to understand their attitudes on the use of wearable devices in educational settings; Conducted open-cod- ing on interview data; Designed survey to understand online students’ attitudes and opinions about the design of wearables and emotion data use

kids making phone call
NatureCollections: Can a mobile app connect kids with nature?
Collaborators: Saba Kawas, Katie Davis
University of Washington Information School

Worked on evaluating the effectiveness of a mobile application in connecting kids with nature. I brainstormed study procedures and designed interview protocol; Led several observational study sessions, conducted focus groups, surveys, and contextual inquiry to gain kids’ opinion on the mobile application; Cleaned and analyzed video data using interaction analysis.

patient provider collaboration
Supporting Patient Provider Collaboration Around Personal Informatics Data
Collaborators: Chia-Fang Chung, Sean Munson
University of Washington HCDE

Worked on exploring the effectiveness of electronic photo-based food diary in supporting patient-pro- vider collaboration. Leveraged the photo-based food journal prototype to understand patient and provider needs; Scheduled and conducted interviews with 17 patient-provider pairs; Analyzed interview data using affinity diagram; Conducted literature review

user satisfaction
User Value Scale
Collaborators: Hyewon Suh, Julie Kientz
University of Washington HCDE

Worked on designing a user value scale to measure users’ perceived value in computing technologies, and how it affect users choices of technology adoption; Brainstormed and defined initial user values through affinity diagram; Designed interview protocols and survey questions; Participated in partici- pant recruitment; Conducted literature review and pilot interviews; Familiarized with Amazon Mechanical Turk(AMT) and helped distribute surveys

pills
Predictors of Sexual Revictimization
Collaborators: Debra Kaysen
University of Washington, Center for the Study of Health and Risky Behaviors (CSHRB)

This project is an individual project advised by Dr. Debra Kaysen and was presented at the 2016 Undergraduate Research Symposium at University of Washington. Women who experienced sexual assault are 2 – 3 times more likely to experience sexual revictimization (SR). Factors that predict risk of SR include emotion dysregulation, guilt, and lower heart rate reactivity when facing sexual threats. These factors overlap with tonic immobility (TI), which is characterized by victim’s diminished or absent physical or vocal response during sexual assault ( “freeze” response).  Post-traumatic maladaptive cognitions also may increase risk for SR, either directly or through increasing risk of TI. Based on prior literature, my hypotheses are: 1. Higher TI will predict higher SV risk; 2. More severe maladaptive cognitions will predict higher SV risk; 3. TI will be predicted by maladaptive cognitions. Research findings suggest that prevention efforts may want to target women who experienced prior TI as a means of preventing future revictimization. These prevention efforts may be more effective if they include cognitive interventions.

NEWS

Travel 6/22/19 In San Diego for the Designing Interactive Systems (DIS) 2019 conferrence
Paper 3/21/19 Our paper Design in the HCI Classroom: Setting a Research Agenda was accepted to DIS 2019 and received best paper award (top 1% of the papers)!!
Paper 1/9/19 Our paper Identifying and Planning for Change: Patient-provider Collaboration Using Lightweight Food DIaries in Healthy Eating and Irritable Bowel Syndrome was accepted to IMWUT with minor revision!!