Home Meeting Notes Student Pages Google Group
COMP.4900/5901 Developing AI Tools for Children, Spring 2023
Prof. Fred Martin,
What Is This?
Developing AI Tools for Children is a directed study course.
We will review the literature on existing work, conceive of and build prototypes of new tools, and then test them with our target audience (middle school-age children).
Please be prepared to do literature reviews, read papers, write code, learn how to do education research, and write a paper.
The Semester's Arc
The semester will involve the following activities and schedule:
Jan through mid-Feb Literature review; how to conduct education research (including human-subjects research training); initial brainstorming of project ideas
second half of Feb Project implementation begins; evaluation design
first half of Mar Project implementation continues; preparation and submission of research proposals to UMLs Institutional Review Board (IRB)
second half of Mar Project implementation concludes
first half of Apr Testing with children
second half of Apr Writing final paper
For the project work (including implementation, research design, data collection, analysis, and writeup), I encourage students to work collaboratively in teams of two or three.
- due Sun Mar 19 per conversation in class, revise your structured interview prompts to three to four questions.
- due Fri Mar 24 per below, make a plan for completion of your project, with three key dates:
- Fri Mar 24: pre-alpha checkpoint: Nearly all features are completed.
- Fri Mar 31: in-class alpha demo: All features are implemented. Bugs are to be anticipated. You will play with each others' systems and note issues.
- Tue Apr 4: beta test of software at Bartlett: Software should be done and debugged. We are testing that everything works in the performance environment.
- Tue Apr 11 and Fri Apr 14: live use with children at the Bartlett School.
- Accomplish the software development work for the pre-alpha checkpoint of your project.
- Su Mo Tu We Th Fr Sa
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 2325 pre-alpha checkpoint 26 27 28 29 30 1 in-class alpha demo April 2023 2 3  5 6 7 8 beta test at Bartlett school 9 1012 1315 live use with children 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
- Create two visual assessments about AI in the style of the MIT PopBots Assessments. Submit into the Assessments folder. Name your files with your name and a few keywords about what it's assessing. These are general assessments (not project-specific).
- Write five short assessment prompts you would like to ask children after they engage with your project. In Project post assessment prompts, make a section for your project (e.g. just its title in bold face) and put the prompts below that header. Each team member should create five potential prompts; put your name above yours. We'll whittle these down for the actual event.
- Your project will accompanied by a trifold posterboard display. Create a sketch of how you will present the activity to children. Be visual. Your posterboard should be mostly images. Post on the
Projectpage of your web site.
- Determine what URLs your system will need to access to be functional. Fill out this spreadsheet with information (a separate row for each URL).
- Continue project implementation work. Notify Fred if you have any change in the core concept you are developing.
- Continue work on projects. Be prepared to show demos at the next meeting.
- Review the study application documents for last spring's AI Afterschool 2022 study led by Ismaila Sanusi. Please read through the following documents:
- HRP-504, main study design
- parent consent
- pre/post survey
- student demographic survey
- promotional descriptions
- Complete the CITI Human Subjects Research training:
- Complete two modules in the only if applicable section: Research with Children and Research in Public Elementary and Secondary Schools. Even if you have the completion certificate, do these two modules if you haven't already.
- Upload your Certificate of Completion to this Google Drive folder. Name your file with this format:
Firstname Lastname citiCompletionReport YYYY-MM-DD.pdf. The date should be the valid until date of your training.
- Team commitments TBD
Readings & Viewing
- Watch Gender Shades by Joy Boulamwini, which addresses bias in face recognition technology.
- Select and read two of the papers chosen by your peers last week. See more in Assignments.
- The state of the art in face recognition has improved since Dr. Boulamwini's research, owing in substantial part to the impact of her work, but the issue of bias in AI technology is still pervasive and embedded. Do a little web-searching on this and describe a few other areas of AI where bias still an issue (~250 words). Post in your Week 3 area of your web site.
- Per above in Readings, choose two of the papers identified by your peers that you'll read. Then, combining with the paper you selected, write a ~500 to 750 word synthesis that compares the three papers. Please focus on papers that involve children in a study. I'm looking for similarities and differences in the research methodologies. Post on your Week 3 page.
- Form a team with one or two other students and develop an idea for your semester project.
- On one team member's web site, create an area where all team members can make edits. Other member(s) of the team should link to the central site from their own sites.
- Write an initial draft of the idea, including:
- overview of the premise;
- which ideas in AI will be taught;
- one or more drawings presenting the how it will look and work (e.g., it could be a storyboard with several frames indicating how users interact)
- a description of how children will interact with the system;
- a discussion of what data you could gather.
- In addition to surveys and interview data, make sure to consider about data that can be gathered electronically as a direct result of children's interaction with your system.
- Make sure to consider how you will engage students in considering the ethics of AI as part of your project.
- If you're working through the CITI course, complete three more modules: Informed Consent, Privacy and Confidentiality, Unanticipated Problems and Reporting Requirements in Social and Behavioral Research. (Two final modules in the only if applicable section will be assigned next week: Research with Children and Research in Public Elementary and Secondary Schools.)
Read the following papers. Reflection prompts are below. Papers can be found in the Google Group.
- Touretzky, D., Gardner-McCune, C., Martin, F., & Seehorn, D. (2019, July). Envisioning AI for K-12: What should every child know about AI?. In Proceedings of the AAAI conference on artificial intelligence (Vol. 33, No. 01, pp. 9795-9799).
- Vaishali Mahipal, Srija Ghosh, Ismaila Temitayo Sanusi, Ruizhe Ma, Joseph E. Gonzales, and Fred G. Martin. 2023. DoodleIt: A Novel Tool and Approach for Teaching How CNNs Perform Image Recognition. In Australasian Computing Education Conference (ACE '23), January 30-February 3, 2023, Melbourne, VIC, Australia. ACM, New York, NY, USA 8 Pages. https://doi.org/10.1145/3576123.3576127
- Williams, R., Park, H. W., & Breazeal, C. (2019, May). A is for artificial intelligence: the impact of artificial intelligence activities on young children's perceptions of robots. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1-11).
- Trivedi, M. R., Movahed, S. V., Perkins, R., Robinette, P., Ahmadzadeh, S. R., & Cabrera, M. E. Behavior Modeling for Robot Theory of Mind.
- Find another paper in the space of AI, Machine Learning, and Data Science tools for K-12 learners. Claim it by posting a citation in the new
#paperschannel in our Discord. Upload it to the Google Drive folder. On your Week 2 web site area, write a 250-summary of the paper. Link directly to your paper from your web site. Make sure your paper hasn't been claimed already when you choose it! (everyone must have a unique paper)
- Think of three concepts in AI/ML/DS you might be interested in developing a project around. Write ~250 words describing each of these three AI concepts and situating each of them in the Five Big Ideas themes developed by Touretsky et al. in the paper above. It's fine if any given of your three concepts fits in multiple themes. (Note: the AI4K12 Big Ideas doesn't really have an area for data science. That's finedon't worry about this if you're interested in a data science idea.)
- For the Williams et al. paper on Pop-Bots, write a ~250-word reflection on how they handled assessing student learning outcomes.
- For the Mahipal et al. and Trivedi et al. papers, write a ~250-word reflection (each) on what is most interesting / surprising about this article. This should not be a summary.
- If you haven't done the IRB Human Subject Training, start it now. Create an account per the directions at the CITI Program Information page. Do the first four modules in the training: History and Ethical Principles, Defining Research with Human Subjects, The Federal Regulations, and Assessing Risk.
Reflections are due on your web sites by 9 AM on Thu Feb 2.
1. Create a web site where you will keep track of your work and make it available to this class for sharing. You can either make it public or make it private but in a way that all of us in this class have access. Send me the link to your web site by 5 PM Tue Jan 24. I will add your site link to this site so we all have access to each other's work. See this class for an example.
2. Read these two papers, available at this OneDrive link:
- Lee, I., Ali, S., Zhang, H., DiPaola, D., & Breazeal, C. (2021, March). Developing Middle School Students' AI Literacy. In Proceedings of the 52nd ACM technical symposium on computer science education (pp. 191-197).
- Long, D., & Magerko, B. (2020, April). What is AI literacy? Competencies and design considerations. In Proceedings of the 2020 CHI conference on human factors in computing systems (pp. 1-16).
Post short reflections (250 words ea) on the papers on your web site. Don't summarize. Instead, focus on what you found surprising or interesting about the papers. Post these to your web site by 5 PM on Wed Jan 25.
3. Create an account at ChatGPT, the new AI chat agent. Use it to get some ideas for ways to engage children in learning about AI. On your web site, share (1) your best ChatGPT prompt and (2) your favorite response from it, due 5 PM on Wed Jan 25.
4. Prior to the next meeting, review the work of at least three of your classmates. Please take private notes.