Team: I worked with a team of user experience researchers, design partners, and product managers. Our team consisted of 6 members working in the U.S.
Duration: I worked on this project for 3 months during my summer internship.
Developed a research plan in collaboration with the design partners, stakeholders, and product managers.
Recruited participants through an external agency and selected appropriate respondents based on the research goals.
Conducted in-depth interviews to understand participants' daily lives as field technicians that visit hospitals for work.
Analyzed data using deductive coding to gain insights into the current work processes and identified their pain points.
Identified opportunities to reduce pain points of healthcare field technicians and mitigate gaps in their workflows.
Communicated the research insights to the interdisciplinary team.
Insights from this study could be used as a basis for future research plans.
----------------------------------------------------------------------------------------------
Team: I worked with a team of user experience researchers, designers, and product managers. Our team consisted of 6 members working in the U.S.
Duration: I worked on this project for 3 months during my summer internship.
Developed a research plan in collaboration with the team.
Recruited participants through an external agency and selected appropriate respondents based on the research goals.
Developed scenarios and tasks for remote usability sessions in collaboration with product managers and designers on the team.
Conducted task-based remote moderated usability testing sessions using a think aloud protocol.
Identified users' pain points and emotions during task completion to gain insights on usability issues in terminology, interactions, and navigation.
Prioritized usability issues based on severity, occurrence, and number of users affected.
Presented research insights to the team and facilitated the prioritization of fixes and new features to implement in future releases.
Insights from this study could be used as a basis for future research plans.
Team: I worked with a collaborative team of human-computer interaction and machine learning researchers, designers, engineers, and clinical subject-matter experts. Our team consisted of 10 members working in the U.K.
Duration: I worked on this project for 3 months during my summer internship.
Performed a thorough literature review on the current state of decision support systems in medical settings and identified gaps in the literature.
Conducted in-depth interviews with decision-makers in the ICU to understand the factors that they consider while making critical decisions in managing patient flow across the hospital.
Analyzed data from ethnographic observations and interviews in order to derive initial explorations of actionable machine learning algorithms.
Identified opportunities and challenges for predictive machine learning models to support the complex decision making processes in clinical contexts.
Communicated results to an interdisciplinary research team to inform and shape the themes for future research.
The complexities of bed management processes and opportunities for machine learning interventions would provide the basis for understanding the intricacies of medical work.
Explorations of possible machine learning interventions would provide an idea of the kinds of problems that are appropriate for these interventions to be actionable in real-world.
This work also contributes towards the design and research considerations for a decision support system that utilizes machine learning in any complex medical context.
My team continues to explore how machine learning could be used to develop actionable predictive analytics in medical settings.
An example narrative schema
Team: I work with a collaborative team of designers, engineers, subject-matter experts, and other researchers. Our team consists of 20 members working in three locations in the North East U.S.
Duration: This work is part of a bigger multi-year project to improve clinical decision-making with the use of technologies in emergency medical scenarios. I am working on this project since September 2016.
Understand the communication behaviors of medical team members in emergency scenarios.
Analyze clinicians' communication patterns to build models of verbal communication (called "narrative schemas") to be leveraged by engineers for system development.
Examine the nature of communication to assess the feasibility of speech as a modality for activity recognition -- its advantages, challenges, and how it could be combined with other modalities.
Performed content analysis on transcriptions derived from audio recordings of emergency medical processes to understand the nature and structure of speech.
Created speech workflow models for 32 activities (Example: Blood Pressure Check, Airway Assessment) that are performed during resuscitation events. These models represent the verbal interactions between medical team members.
Conducted ethnographic fieldwork including in-situ observations during resuscitation events to understand the communication and information sharing among clinicians.
The speech workflow models that we constructed would be used in decision-support system development by the engineering team to facilitate speech-based automatic activity recognition.
This work also contributes towards design and research considerations for an activity recognition system in any multi-user, speech-heavy environment.
We plan to build heuristics to combine speech workflow models with neural networks to facilitate activity recognition.
We continue to explore how speech could be combined with other modalities and how those modalities could overcome the challenges faced by speech-based activity recognition systems.
Methods used: content analysis, discourse analysis, ethnography, participant observation
Publications (links available on the Publications page)
[1] Jagannath, Swathi, Aleksandra Sarcevic, Neha Kamireddi and Ivan Marsic (2019). Assessing the feasibility of speech-based activity recognition in dynamic medical settings. CHI ’19, Proceedings of the ACM SIGCHI Conference on Human Factors in Computing Systems: Extended Abstracts, Glasgow, UK, May 4 – 9, 2019.
[2] Jagannath, Swathi, Aleksandra Sarcevic and Ivan Marsic (2018). An analysis of speech as a modality for activity recognition during complex medical teamwork. PervasiveHealth ’18, Proceedings of the 12th EAI International Conference on Pervasive Computing Technologies for Healthcare, New York City, NY, USA, May 21 – 24, 2018.
[3] Jagannath, Swathi and Aleksandra Sarcevic (2018). Speech modeling for automatic activity recognition in complex medical settings. PHICWIC ‘18, Presentation at ACM Philadelphia Region Celebration of Women in Computing, Philadelphia, PA, USA, April 20 – 21, 2018.
[4] Jagannath, Swathi and Aleksandra Sarcevic (2017). Using speech in complex teamwork to support activity recognition. WISH ‘17, Proceedings of the 2017 Workshop on Interactive Systems in Healthcare, Denver, CO, USA, May 7, 2017.
Resuscitation bay at our research site, where we perform in-situ participant observations during patient care.
Team: I work with a cross-functional team of designers, developers, and subject-matter experts. Our team consists of six members working in two sites (hospital and university).
Duration: This work is part of a multi-year project to improve documentation in emergency medical scenarios. I started working on this project in April 2017.
Examine nurses' documentation and communication behaviors on a newly implemented electronic flowsheet associated with the EHR.
Investigate user attitudes with electronic documentation to gain insights into challenges with EHR use during resuscitation events.
Examined the nuances of the transition process from paper to electronic documentation system during fast-paced medical scenarios.
Conducted ethnographic field work (766 hours thus far) including participant observations and semi-structured interviews with the clinicians.
Reviewed video recordings and EHR logs to gain insights into the documentation patterns on the flowsheet.
Identified patterns such as early, on-time, and delayed documentation on the EHR sections by comparing the charting times to the actual activity performance times on the video recordings.
Investigated the use of paper-based workarounds during electronic documentation by collecting the paper notes used by clinicians to document during resuscitation events.
The temporal patterns of documentation that we have identified are used to provide design and research implications to designers and other researchers in order to improve the usability and thus support real-time documentation on the EHR.
Our findings on paper-based workarounds contributes towards supporting the work practices of nurses by providing directions to align the design of electronic documentation tools with their actual workflows.
We continue to investigate the communication behaviors of team members to identify temporal aspects communicated via speech.
Temporal aspects of speech could be used to design a display to alert teams on process delays.
Methods: participant observations (in-situ and video review), semi-structured interviews, contextual inquiries, log analysis, artifact collection and analysis, affinity diagramming
We found that the nurses heavily take notes on paper during dynamic medical scenarios, thus affecting real-time documentation on EHRs
Publications (links available on the Publications page)
[1] Jagannath, Swathi, Aleksandra Sarcevic, Victoria Young and Sage Myers (2019). Temporal rhythms and patterns of electronic documentation in time-critical medical work. CHI ’19, Proceedings of the ACM SIGCHI Conference on Human Factors in Computing Systems, Glasgow, UK, May 4 – 9, 2019.
[2] Jagannath, Swathi, Aleksandra Sarcevic and Sage Myers (2019). From paper forms to electronic flowsheets: Documenting medical resuscitations in a time of transition. Proceedings of iConference 2019, Washington D.C., USA, March 31 – April 1, 2019.
[3] Jagannath, Swathi, Aleksandra Sarcevic, Nina Multak and Sage Myers (2018). Understanding paper-based documentation practices in medical resuscitations to inform the design of electronic documentation tools. Pediatric Emergency Care, 2018.
[4] Jagannath, Swathi, Aleksandra Sarcevic and Andrea Forte (2018). “We are not entirely replacing paper”: Understanding paper persistence in emergency medical settings. CSCW ’18, Proceedings of the 21st ACM Conference on Computer-Supported Cooperative Work and Social Computing: Extended Abstracts, Jersey City, NJ, USA, November 3 – 7, 2018.
[5] Jagannath, Swathi (2018). Transition from paper to electronic documentation in pediatric emergency medical settings. GHC ‘18, Poster Presentation at Grace Hopper Celebration of Women in Computing, Houston, TX, USA, September 26 – 28, 2018.
Team: We are a team of three graduate students with diverse backgrounds in Human-Computer Interaction (HCI), Data Analytics, and System Modeling.
Duration: This was originally submitted as a class project in December 2017. We planned to continue working on it in November 2018.
Examine the technology considerations taken by international students when they plan to migrate to a developed country to gain insights on the differences in technology diffusion in different countries.
Investigate the current resources available for international students to support the technology transition and develop guidelines for addressing any gaps in resource availability.
Recruited international students to participate in the study by answering survey.
Developing a follow-up interview to gain a deeper understanding based on survey responses.
This is a new, on-going study.
Methods: surveys, interviews
Publications (links available on the Publications page)
[1] Kimberley Hemmings-Jarrett, Swathi Jagannath, Ali Jazayeri, and Denise Agosto (2019). “We Need More Than Laptops!” Technology Assistance for Transitioning International Students. CSCW ’19. Proceedings of the 22nd ACM Conference on Computer-Supported Cooperative Work and Social Computing: Extended Abstracts, Austin, TX, USA, November 11 – 13, 2019.
[2] Jagannath, Swathi, Kimberley Hemmings-Jarrett and Ali Jazayeri (2018). Understanding information and communication technology diffusion in developing countries. AMCIS ’18, TREO Proceedings of the 24th Americas Conference on Information Systems - Digital Disruption, New Orleans, LA, USA, August 16 – 18, 2018.
Team: I worked with a team of developers and other researchers co-located in our university.
Duration: This work was part of a bigger project. My contribution was between January 2012 - June 2013.
Examine errors that people make and their cognitive sources to provide design recommendations for the development of a web editor in order to support novices' development practices.
Understand how novices recover from the errors they made to gain insights on how those errors supported learning and how these learning practices could be incorporated in the web editor.
Conducted usability studies in a lab setting where the participants completed a set of HTML and CSS coding tasks using think-aloud protocol.
Performed open and axial codings of errors made by participants using the video recordings of the study sessions.
Developed a taxonomy of errors using the skills-rules-knowledge framework to understand when and why the errors occur in order to provide learners with the means to detect, understand, and resolve them productively.
Derived design implications for a translational web editor by understanding the error making and resolving patterns of novice web developers.
Our team continues to work on understanding learning trajectories with web development to develop better tools to support those trajectories.
Methods: usability studies, contextual inquiries
Details on this project could be found here.
Publications (links available on the Publications page)
[1] Park, Thomas, Ankur Saxena, Swathi Jagannath, Susan Wiedenbeck and Andrea Forte (2013). Towards a taxonomy of errors in HTML and CSS. ICER ‘13, Proceedings of the Ninth Annual International ACM Conference on International Computing Education Research, San Diego, CA, USA, August 12 – 13, 2013. pp. 75-82.
[2] Park, Thomas, Ankur Saxena, Swathi Jagannath, Susan Wiedenbeck and Andrea Forte (2013). openHTML: Designing a transitional web editor for novices. CHI ‘13, Proceedings of the ACM SIGCHI Conference on Human Factors in Computing Systems: Extended Abstracts, Paris, France, April 27 – May 2, 2013. pp. 1863-1868.
Team: I worked with a team of two graduate students with Information Science background.
Duration: This was a class project (12 weeks).
Understand how people use social media to share information during emergencies to incorporate and facilitate technology support from government and federal agencies.
Examine the patterns of information seeking and sharing on social network to derive design implications in order to handle the spread of rumors and misinformation in critical emergency situations.
Investigate how users connect with their family and friends to make them aware of their personal situation and safety and thus define directions to provide technology support in cases of infrastructure breakdowns.
Gathered requirements by conducting short interviews with users that have been in emergencies and also a user experience researcher who has a background in crisis management.
Created personas and scenarios to outline the goals, behavior, attitudes, and motivations of target users.
Conducted cognitive walkthroughs using high-fidelity prototypes to understand the learnability of the system.
Identified three dimensions of information flow and barriers during emergency situations -- communication issues due to infrastructure failure, spreading of rumors on social networks and the lack of online participation of government agencies.
Developed three design prototypes to address the major barriers regarding communication and information sharing during emergencies.
Conduct in-depth interviews with users who have been in emergency situations to better understand their information needs.
Perform usability testing on a working prototype to understand how users use the system and thus improve the design.
Methods: requirements gathering, stakeholder interviews, personas, scenarios, prototyping, cognitive walkthroughs, contextual inquiries
Details on this project could be found here.