Syllabus

Course description

CS-770 is a graduate seminar and research-methods course in Human-Computer Interaction, cross-listed across Computer Sciences, Psychology, and Educational Psychology. It has three interwoven strands: a reading seminar surveying the landscape of HCI, a research-methods sequence, and a semester-long team research project that culminates in a conference-style paper and a final showcase.

This year the course treats artificial intelligence — especially generative and agentic AI — as a transforming force across HCI, without losing the enduring fundamentals. Our organizing idea: fundamentals are the medium, not the message. Each week we ask what the durable concern of an area is, how AI is changing it, and — honestly — how much. The fundamentals are the instrument you use to read the AI moment, not a relic to move past.

Learning goals

By the end of the course you will be able to:

1. Define research questions, construct hypotheses, map the literature, and situate questions in existing knowledge.
2. Recognize seminal and contemporary research across the breadth of HCI.
3. Choose a research approach that fits a question; identify variables; design qualitative, quantitative, and hybrid studies.
4. Select appropriate objective, behavioral, physiological, subjective, and composite measures.
5. Design survey items, construct scales, and assess reliability and validity.
6. Analyze qualitative and quantitative data (grounded theory; descriptive & inferential statistics, including regression and effect sizes).
7. Carry out an original HCI research project as part of a team.
8. Write and present an academic paper reporting your design and findings.
9. Critically assess how AI transforms — and does not transform — the fundamentals of HCI, and evaluate AI-infused systems with appropriate rigor.

How the course works

The two weekly sessions are decoupled by strand:

• Mondays — Seminar (topic of the week). Everyone posts a one-slide contribution on the readings before class. In class: instructor synthesis of the topic, then table discussion in your standing pod with report-outs.
• Fridays — Research Methods. A method per week (lecture + hands-on activity), read against the textbook and applied to your project. Two Fridays are studio crits.
• Project (all term). Teams of 3–4 pursue an original HCI research question, with milestones from topic selection through a final paper (ACM acmart sigconf two-column format, Overleaf) and a final showcase.

Studio crits. Your standing pod critiques each team’s work at two milestones — method design (mid-Oct) and analysis/results (mid-Nov) — plus a capstone studio before presentations. Crits are formative (for learning) and run by the pods themselves. Final presentations (Dec 7) run as lightning talks with light peer-voted awards.

Research ethics & IRB for the course project. Your 770 project is a training activity and does not require IRB approval — that is what makes data collection possible inside a 14-week term. The ethical obligations are unchanged: informed consent, proportionate risk, careful data handling. But training data is not publishable data. IRB approval cannot be granted retroactively, so if you want to publish work from your project you must either (1) obtain IRB approval yourself, before collecting the data you intend to publish — I cannot file on your behalf — or (2) treat the class project as a pilot and recollect the data under an approved protocol after the course. If your project looks publishable, raise it with me early.

Required materials

• Textbook: Lazar, Feng & Hochheiser, Research Methods in Human-Computer Interaction, 2nd ed. (Morgan Kaufmann, 2017) — free via the UW-Madison Libraries.
• Weekly readings: seminal + 2024–26 papers per topic, in three tiers — required (one or two per session), recommended, and optional. Listed on the Schedule & Readings page and on Canvas.
• Tools: Canvas, Overleaf, the course website, and statistical software (R or JMP).

Grading (100 points)

Letter grades: A 93.5–100 · AB 89.5–93.4 · B 83.5–89.4 · BC 79.5–83.4 · C 73.5–79.4 · D 63.5–73.4 · F < 63.5.

Teamwork

The project is a team of three, and a team of three is where free-riding is easiest and most corrosive. Six of the forty project points are teamwork, assessed by confidential peer evaluation twice — mid-semester (Oct 26) and end of term (Dec 7). You rate each teammate, and yourself, on doing the work, reliability, and collaboration, plus one sentence on what that person contributed that the team could not have done without.

Your score is your own teamwork mark, derived from what your teammates report — not from the ratings you give. Submitting is required; not submitting scores zero. Ratings are confidential and are never shown to your team. They are used to find genuine imbalance, not to rank people — a well-functioning team of three should all score at or near full marks.

Attendance & participation

Ten points, in three parts.

Attendance (5). Taken by the TA via a quick check against your assigned pod seat — no sheet to pass around. Everyone gets 2 free absences (no questions asked); documented emergencies beyond that are excused. The course meets 26 times, so past your two free absences each further one costs about a fifth of a point.

Pod role (3). Each Monday your standing pod fills three rotating roles — Reporter (writes the pod’s three-line summary and delivers it to the room), Skeptic (voices one counterargument, whether or not you hold it), Connector (links this week to a previous one). In a six-person pod across eleven seminars you take a role in five of them and cycle through all three roles at least once. The rotation is published in advance, so you always know yours. Credit is for fulfilling the role when it comes round — not for how much you talk.

Participation self-assessment (2). Two half-page reflections — mid-term (Oct 26) and end of term (Dec 7) — making the case for your own contribution with evidence. Contribution here is not only verbal: written crit feedback, pod work, and report-outs all count, and this is where you say so.

Because pod roles require being in the room, sustained absence costs you twice — that is by design.

Late work

Each student has a bank of 3 late days to spend across individual deliverables (no penalty). Late days cover seminar posts and methods assignments; project deliverables can’t use them. The final paper and presentation cannot be moved with late days. Beyond the bank, late work is penalized 10%/day.

Collaboration on methods assignments

Methods assignments are individual, but you don’t have to work alone. Talk through the method with classmates, trade feedback on drafts, and take part in each other’s pilot studies — that’s encouraged. What you submit must be your own work: your own choices, reasoning, and writing. Two submissions should never read alike, and co-written or shared write-ups aren’t accepted. Name anyone you worked with at the top of your submission and say how (e.g., “Discussed study design with A. Rivera; piloted with J. Park’s team”).

Use of AI tools

AI is both a subject and a tool of this course, and you are expected to use it. Use it to do your best work — not to cut corners. The distinction that matters is not which tool you touched; it is whether the intellectual contribution (research questions, study design, analysis, argument) is yours and your team’s.

• No disclosure required for coursework. Don’t write disclosure notes on weekly assignments — that’s not a burden I want to put on you.
• Mind the jagged frontier. These systems are excellent at some tasks and confidently wrong at neighbouring ones, and the boundary is invisible and moves. In a field experiment with 758 consultants, AI raised quality by over 30% on tasks inside the frontier — and cut correctness by 19 percentage points on a task outside it (Dell’Acqua et al., 2026). Learning where that edge falls for your work is part of what you’re here to develop.
• You own what you submit. Every claim, number, and citation is yours to stand behind. Fabricated data, results, or citations are academic misconduct regardless of what produced them.
• Venue rules change, and they override this one. Journals and conferences set their own requirements. Our final paper follows ACM’s: generative-AI use is disclosed in a brief Acknowledgements note.
• Ask me. If you want help building good practice with these tools, come to office hours — that’s a conversation I’m glad to have.

We will also, throughout the course, study how to evaluate and design with these systems — so treat your own use as a case study.

Communication

• Course-content questions → in class, email, or office hours.
• Personal / logistical → email the instructional team at hci-class@cs.wisc.edu.
• Feedback on your work → office hours.

Course credit

This is a 3-credit course meeting for two 75-minute periods per week. Consistent with the UW-Madison credit-hour policy (at least 45 hours of learning per credit), you should expect to spend roughly 9 hours per week total on the course — the ~2.5 hours in class plus about 6–7 hours of reading, project work, analysis, and writing outside class. (Policy: UW-1011, The Credit Hour.)

University policies & resources

This syllabus is authoritative for the following University of Wisconsin–Madison policies; each links to the official campus page. (Standard statements per UW-1022; canonical set at CTLM Syllabus Statements.)

• Academic integrity. By enrolling you agree to uphold UW-Madison’s academic standards. Academic misconduct may result in sanctions ranging from a failing grade to expulsion. → conduct.students.wisc.edu/academic-misconduct (UWS 14)
• Disability accommodations. UW-Madison provides reasonable accommodations to students with disabilities. If eligible, please share your McBurney VISA early in the semester; disability information is confidential under FERPA. → mcburney.wisc.edu (UW-855)
• Academic calendar & religious observances. Notify the instructor within the first two weeks of any need for flexibility due to religious observances (UW-880). → secfac.wisc.edu/academic-calendar
• Student health, well-being & basic needs. Help is available for stressors outside class. → students.wisc.edu/guides/get-help-now · University Health Services
• Sexual misconduct & mandatory reporting. UW-Madison prohibits sexual harassment, sexual assault, dating and domestic violence, and stalking (UW-146). The instructor and TA are Responsible Employees; confidential resources are available. → compliance.wisc.edu/titleix
• Students’ rules, rights & responsibilities — including privacy rights and the complaint / grievance process. → guide.wisc.edu
• Privacy of student records & recorded lectures (FERPA). Course materials and recordings are protected; use them for personal, course-related study only. → registrar.wisc.edu/ferpa
• Course evaluations. Near the end of term you will be invited to complete a confidential digital course evaluation. → assessment.wisc.edu/course-evaluations
• Teaching & learning data transparency. UW-Madison vets the digital tools used for teaching and learning to protect your information. → teachlearn.provost.wisc.edu

Schedule (calendar weeks)

Full detail (readings, activities, milestones) is on Canvas; topics and readings are on the Schedule & Readings page.

Final paper due Mon Dec 14, 11:59 pm (finals week). Schedule subject to minor adjustment; changes announced on Canvas.