User Experience and Interactive Systems
In industry, UX work has shifted from specifying deterministic screen behavior to designing around systems that generate probabilistic output, so practitioners now own disclosure, uncertainty display, correction paths, trust calibration, and the escalation points where an agent hands control back to a person. The course keeps its HCI identity and its research-prototype-test cycle, but the artifacts under study now include conversational and agentic features, and evaluation now has to measure recovery from confidently wrong machine output.
Current catalog prerequisites — (CSC 200 or 2000).
Current description → proposed description
This course concerns the fundamental issue of effective and usable human computer interaction. In addition to technical issues, people and process must be understood to create effective and usable tools. As CS and IT practitioners create and manage systems as effective problem-solving tools for others, they must develop a user-centered perspective within the organizational context. To that end this course will study related issues including cognitive principles, human-centered design, ergonomics, accessibility, emerging technologies and usable environments. CSC 3020 is part of the AI concentration in the CS curriculum.
Prerequisites: (CSC 200 or 2000).
This course concerns the design and evaluation of effective, usable, and accessible human-computer interaction, extending human-centered design practice to interfaces whose behavior is generated rather than fully specified. Students study cognitive principles, interaction design patterns, ergonomics, accessibility standards, and usability evaluation methods within an organizational context, then apply them to conversational, assistive, and agentic features whose output is probabilistic and occasionally wrong. Topics include user research and task analysis, prototyping and iterative critique, designing for uncertainty and graceful error recovery, disclosure and provenance of machine-generated content, trust calibration, reversible actions, and human-in-the-loop checkpoints at which an autonomous system must return control to a person. Students conduct usability tests with representative participants, including users of assistive technology, and examine manipulative design patterns and undisclosed automation in light of a Christian understanding of human dignity. CSC 3020 remains part of the AI concentration in the CS curriculum.
What changes
- Designing for probabilistic, machine-generated interface behavior
- Agentic interaction patterns: disclosure, uncertainty, reversibility
- Human-in-the-loop escalation and trust calibration
- Usability testing extended to error recovery from wrong AI output
- Explicit critique of dark patterns and undisclosed automation
6 proposed outcomes, mapped to 12 program outcomes
Each outcome below is written to be observable and assessable, and each is mapped to the program learning outcomes for which it produces evidence.
Students will be able to analyze user research evidence, including contextual observation, interviews, and task analyses, to specify interaction requirements for a defined user population and context of use, including features whose output is machine-generated.
Deriving interaction requirements for machine-generated features from field observation and interviews applies communication and social-science method to the development of a human-centered AI application, which is exactly what PLO 3.2 names.
Students will be able to construct interactive prototypes, including AI-assisted features, that apply cognitive principles, established interaction patterns, and recognized accessibility standards for keyboard, screen-reader, and low-bandwidth use.
Building AI-assisted features that still work for keyboard, screen-reader, and low-bandwidth users is a concrete act of compassion toward people the default interface would exclude, which is the service-through-building standard of PLO 2.1; and making those same AI features reachable on assistive technology and constrained connections is directly the equitable access to AI technologies required by PLO 2.3.
Students will be able to design interaction patterns for AI-assisted and agentic features, including disclosure of machine-generated output, communication of model uncertainty, reversible actions, and human-in-the-loop escalation.
Proposing and defending a specific escalation and undo design for an autonomous feature is the evaluation of an AI-driven solution called for in PLO 4.2, while rendering confidence and provenance legible to an ordinary end user is precisely the translation of technical AI knowledge for non-specialists in PLO 5.3.
Students will be able to evaluate an interactive system through moderated usability testing, accessibility audit, and review of interaction telemetry, including task success and recovery when AI-generated output is wrong, interpreting the resulting evidence to prioritize design revisions.
Interpreting telemetry on how users fared when AI-generated output was wrong, and explaining that result to a design team, is the interpretation and explanation of AI data outcomes described in PLO 6.1; and ranking revisions from task-success and recovery rates alongside observations from moderated sessions and the accessibility audit is the paired statistical and qualitative analysis of PLO 6.2.
Students will be able to critique interface designs for manipulative patterns, undisclosed automation, and inequitable outcomes, defending non-coercive alternatives grounded in the conviction that every user bears God's image.
Arguing that a user is owed disclosure and an honest exit is the transparency-and-accountability application of Christian values in PLO 1.2; examining how an engagement-optimized design harms particular users is the ethically-framed analysis of data-driven outcomes in PLO 4.1; and judging whether a design treats groups of users fairly is the moral assessment required by PLO 6.3.
Students will be able to articulate design rationale, tradeoffs, and evaluation findings for AI-assisted interface features to both technical and non-technical stakeholders in written documentation and a live design critique conducted within a design team, incorporating peer critique into a revised rationale.
Documenting and defending an AI feature's design rationale aloud without overstating what the evaluation showed is the transparent written and spoken communication of AI concepts in PLO 5.1, and conducting the critique within a design team and revising the rationale in response to peers' differing judgments is the teamwork and navigation of diverse views in PLO 5.2.
Program outcomes this course reaches
Filled cells are program learning outcomes with at least one supporting course learning outcome in this course. Sparse coverage is expected — no single course carries all eighteen.