Week 12: Best of CHI 2021

What Do Hackathons Do? Understanding Participation in Hackathons Through Program Theory Analysis

• Falk, J., Kannabiran, G., & Hansen, N. B. (2021, May). What Do Hackathons Do? Understanding Participation in Hackathons Through Program Theory Analysis. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems.

• Hackathons
  • Put on by organizers
    • Makerspace
    • Hackerspace
    • etc…
  • not just for nerds any more.
  • rapid prototyping resource for companies
• Program Theory
  • approach used to understand
    • processes
    • how to evaluate a program
  • 
  • inputs
    • resources needed to initiate and complete a program
      • eg.
        • design materials or tools in hackathon
        • design challenge
  • process
    • activities that are conducted during program by participants
      • mechanisms are supported by activies
        • general principles that generate effects.
        • eg.
          • mid-session hackathon workshops to improve prototyping proficiency
            • could lead to better quality prototypes
  • effects
    • outputs
      • immediate tangible and intangible products of process
      • eg.
        • prototypes made during hackathon
    • outcomes
      • short and midterm
      • eg.
        • consequences, benefits or drawbacks of program
        • startup company made from hackathon idea
    • impacts
      • long term
      • hard to determine specific hackathon cause
      • eg.
        • hope and belonging by members
• Method of Analysis of papers
  • hackathon domain (edu, research, civic)
  • main organization approach for specific hackathon (choice of material)
  • reason for format modification
  • elements
    • input
    • process
    • output

“Can I Not Be Suicidal on a Sunday?”: Understanding Technology-Mediated Pathways to Mental Health Support.

• Pendse, S. R., Sharma, A., Vashistha, A., De Choudhury, M., & Kumar, N. (2021, May). “Can I Not Be Suicidal on a Sunday?”: Understanding Technology-Mediated Pathways to Mental Health Support. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems.

• Empirical study
• Phone hotlines in India
  • barriers for care
    • societal norms
      • marginalization
      • stigma
      • depression
    • personal identity
    • laws
    • institution aversion
• Pathways to Care model
  • Goldberg and Huxley
  • maps process for assigning mental illness to people
  • power differential problems

“Why lose control?” A Study of Freelancers’ Experiences with Gig Economy Platforms.

• Carlos Alvarez de la Vega, J., E. Cecchinato, M., & Rooksby, J. (2021, May). “Why lose control?” A Study of Freelancers’ Experiences with Gig Economy Platforms. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems.

• Empirical study
• Qualitative Analysis of four Reddit subreddits
  • Showed tensions between
    • wanting autonomy and control
    • necessity of opportunities and convenience

Understanding data accessibility for people with intellectual and developmental disabilities.

• Wu, K., Petersen, E., Ahmad, T., Burlinson, D., Tanis, S., & Szafir, D. A. (2021, May). Understanding data accessibility for people with intellectual and developmental disabilities. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-16).

• Analysis

  • 
  • measured performance for different visualization designs
    • four tasks
  • interviews
  • independent vars
    • chart type
    • chart embellishment
    • data continuity
    • ability level
  • dependent vars
    • task completion accuracy
    • response time

Heuristic Evaluation of Conversational Agents.

• Langevin, R., Lordon, R. J., Avrahami, T., Cowan, B. R., Hirsch, T., & Hsieh, G. (2021, May). Heuristic Evaluation of Conversational Agents. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems.

• Heuristic evaluation type
• Evaluated metrics for testing robot performance for two interfaces:
  • chatbot
  • voice
• The four phased design process.
  • Phase 1: Heuristic Generation
  • Phase 2: Expert Review
    • 
  • Phase 3: Validation through Heuristic Evaluation
    • The conversational agent heuristics developed in Phase 3
      • Visibility of system status
      • Match between system and the real world
      • User control and freedom
      • Consistency and standards
      • Error prevention
      • Help and guidance
      • Flexibility and efciency of use
      • Aesthetic, minimalist and engaging design
      • Help users recognize, diagnose and recover from errors
      • Context preservation
      • Trustworthiness
  • Phase 4: Validation of Revised Heuristics
    • 
• Better evaluators than Nielsen’s.
  • Identified issues related to
    • dialog content
    • interaction design
    • help and guidace,
    • human-like cyharacteristics
    • data privacy

“That courage to encourage”: Participation and Aspirations in Chat-based Peer Support for Youth Living with HIV.

• Karusala, N., Seeh, D. O., Mugo, C., Guthrie, B., Moreno, M. A., John-Stewart, G., Inwani, I., Anderson, R. & Ronen, K. (2021, May). “That courage to encourage”: Participation and Aspirations in Chat-based Peer Support for Youth Living with HIV. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems.

• Qualitative and Empirical study (Kenya)
• Stigma with HIV causes privacy issues
  • phone sharing
• Data collection: surveys, notes, chat records, and audo recorded interviews.
• The Rewards and Limitations of Facilitated Peer Support
  • Exchanging Emotional and Informational Support
  • Rediscovering Aspirations and Ways to Achieve Them
  • Feeling the Limitations of Peer Support
• Problems:
  • High volume of unstructured data
  • Chat group can be taken over by non-facilitators
• Positive Results:
  • found that aspirations, along with a sense of community, could motivate participants to try to stick with difcult tasks.

Designing Interactive Transfer Learning Tools for ML Non-Experts.

• Mishra, S., & Rzeszotarski, J. M. (2021, May). Designing Interactive Transfer Learning Tools for ML Non-Experts. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-15).

• Transfer Learning
  • Reuse trained ML models (transfer to new application)
  • Still mostly experts-only
    • curated (model zoos) domain-specific pre-trained toolkits
    • lots of subtasks
      • identify transfer candidates
      • successful transfer
      • performance eval
      • next step identification

• Study of non-expert behavior during transfer learning activities

  • focus on convolutional neural nets
    • identify object boundaries, colors and shapes in an image
    • transfer those to models
    • filters exposed as “layer” with features
      • “layer” would match novice user intuition better

• participants

  • self-selected via:
    • knowledge of reading and writing about/for a given domain
    • desire to learn about ML to complete a task (side-benefit)

• Empirical (Qualitative) observations

  • 

• Conceptual Model

  •