Design
Human-Computer Interaction
17 Aug 2026
Week EIGHT
Part VI of Hornbæk et al. (2025)
Intro
Everything you interact with was designed. Someone decided which functions to include, what the icons look like, and how the pieces fit together. When something “just works” we forget that a great many decisions went into making it that way—and when it is cumbersome, we feel every one of them.
Why design matters
- Design affects how we perceive, experience, remember, decide, and form habits
- Small improvements accumulate into large effects across millions of uses
- Design is a central driver of innovation and business—much of today’s revenue comes from products and services that did not exist five years ago
- Design is more than making things look nice; it pursues exploratory, critical, and artistic goals
Part VI at a glance
This part covers four chapters
- Introduction to design (Ch 30): what design is
- Design cognition (Ch 31): how designers think
- Design practice (Ch 32): how designers actually work
- Design processes (Ch 33): how design is organized
Introduction to design (Ch 30)
- Design is not the privilege of professional designers
- Hornbæk et al. (2025) quotes Herbert Simon: “Everyone designs who devises courses of action aimed at changing existing situations into preferred ones”
- The challenge here is to use our knowledge of people (Part II), needs (Part III), and interfaces (Parts IV–V) to design new interactive systems
Four views of design
Design means at least four different things
- Design as product: the artifact or plan that results—the ultimate particular
- Design as process: the activity of designing, a form of problem-solving
- Research through design: the generation of new knowledge, not just artifacts
- Design as change: design’s transformative, even political, potential
Design as product
- The outcome is an ultimate particular: a concept, plan, or artifact that realizes an idea in a testable form
- Sketches and wireframes count—they materialize the designer’s imagination so it can be assessed
- Gaver: ultimate particulars are about possible worlds
- Natural sciences ask how things are; design asks how things ought to be
Design as change
- Design is not only about objects you can hold—it can shake the world and get people to think and act in new, better ways
- Buchanan (1992) saw design as a means of changing culture and human experience
- Instead of merely fixing problems (solutionism), design can create transformative possibilities
- Example: the Nintendo Wii transformed a video console into a device for physical activity
- The catch: this makes designers’ thinking political—they must take a stance on what is desirable
Aside: research through design
- Curious Cycles is a research-through-design project on the experience of menstruation
- It used a design probe: a kit sent into the wild for participants to engage with, without the researcher present
- Probes purposefully exploit ambiguity—participants interpret their own experiences
- An unexpected finding: participants shared materials with each other, improving self-knowledge
Interaction design
- HCI’s slice of the design world is interaction design, defined by two characteristics
- a focus on interactive technology
- human-centeredness
- Designing a secure payment protocol is not HCI; designing the interface for those payments is
- The point of a design is realized only in the context where it is used
The designer’s fundamental uncertainty
- Simon distinguished the inner environment (how the artifact behaves) from the outer environment (where it is used)
- We can only design the inner environment—and hope it interacts well with the outer one
- So a design is, in a sense, a hypothesis about that interplay
- No matter how well the artifact is controlled, its potential is uncertain until it meets real use—only evaluation and deployment reduce that uncertainty
A simplified view of design—the designer accrues design knowledge as artifacts meet users across many contexts.
Four rationales for design in HCI
- Improving: design can improve usability, accessibility, and experience
- Creating: design can open entirely new opportunities for using computers
- Informing: by showing alternatives to the status quo, design informs decisions and creates the need for change
- Producing knowledge: design does not just apply HCI knowledge, it evolves it
User-centered design
- Human-centeredness: the primary aim of design is to improve human conduct
- User-centeredness: “The needs of the users should dominate the design of the interface”
- Distilled into a handful of principles
Principles of user-centered design
- User focus: serve users’ goals, tasks, and needs; avoid jargon
- User involvement: engage representative users throughout
- Iterative and incremental development: you cannot know the specifics up front
- Prototyping: evaluate ideas before converging
- Evaluation in context: test with real users in real settings
- Holistic design: everything from ads to manuals affects use
- Process customization: avoid rigidity; keep reflecting
Generating creative ideas
- Design is hard because creative ideas are hard—novelty alone is not enough, an idea must also meet objectives
- The design space is enormous: choosing which of \(n\) functions to include gives \(2^n-1\) options
- For just 50 functions, that is over \(1.1\times10^{15}\) possibilities—before you even organize them into a menu (\(50! \approx 10^{64}\) arrangements)
- Design tasks are also often poorly defined: “design a better version” invites the questions better for whom and in what sense
Boden’s three types of creativity
- Combinatorial: a novel combination within a known conceptual space (the Dvorak keyboard optimized letter placement)
- Exploratory: a new conceptual approach revealing previously unthinkable ideas (ubiquitous computing)
- Transformative: replacing a conceptual space entirely (the desktop metaphor)
Design thinking
- A popular, emancipatory idea: creativity is not the gift of a “design hero”—it is something anyone can develop
- Offers methods for divergent thinking (chart the options) and convergent thinking (narrow to the best)
- Emphasizes two activities: knowing what to design and knowing how to design it
- Knowing what to design is arguably the harder and more important part
Aside: the double diamond
The Design Council’s model has two diamonds
- First diamond—Discover then Define: understand and frame the problem
- Second diamond—Develop then Deliver: build and refine the answer
- Each diamond opens with divergence and closes with convergence
- Four principles: put people first, communicate visually, collaborate, and iterate
The double diamond: diverge then converge, twice—first on the problem, then on the solution.
Practicing design
- Creativity rarely happens spontaneously—it follows from deliberate technique and practice
- Design techniques: defined ways of performing design tasks
- Design practices: how designers actually think and work, reflecting a designer’s maturity
The reflective practitioner
- Schön (1983) argued that professional designers develop distinctive ways of knowing through practice—they reflect on their creations
- Real design problems lack the tidy goals of engineering; design is messy
- Buchanan: designers contextualize their work through placements—cycles that let them make sense of what a design is for
- These meta-cognitive skills are the product of training, not talent or DNA
Wicked problems
- A wicked problem is hopelessly constrained or underdefined
- “Innovate something for an automobile company”—how would you even begin?
- Wicked problems have no stopping rule and no clear goal
- Some even have indeterminacy: no definitive boundaries at all
- Reflection and contextual placement let a designer iteratively crack them
Practices of participation
- Participatory design: stakeholders help generate ideas, not just get studied—born in 1970s Scandinavia to democratize the workplace
- The researcher’s role shifts from translator to facilitator
- Co-design: the whole process is rethought as a collective effort; authorship is shared with user communities
- Action research: to change the world, the designer must become part of it—and accept public accountability
What should be designed?
- The key question in human-centered design is not what can be designed but what should be
- Value-sensitive design: build designs rooted in principles users find important
- A startling implication: the designer must be ready to not design when the data suggest negative outcomes for users
- Human-centricity is empirical—it is grounded in user research from early stages to evaluation
Aside: value-sensitive design of AI
- Intelligent systems are often built algorithm-first: grab a dataset, train a model, report accuracy
- But an algorithm deciding insurance policies, tumors, or job resumes lives inside a system
- Zhu et al. applied value-sensitive design to Wikipedia edit-war communities in five steps: understand the people, create prototypes, elaborate methods for working with users, learn from deployment, and assess outcomes
- The lesson: situate an algorithm as part of a system, not merely in relation to data
Summary of Ch 30
- Design reaches beyond the user interface to graphics, concepts, and services
- Design is about changing users’ practices and experiences—natural sciences seek knowledge but do not aim to change the world
- Designers’ work is organized into processes involving techniques and practices
- Design is the nexus of human-centered design where user research, evaluation, and engineering meet
Design cognition (Ch 31)
- Try it: spend ten minutes sketching alternative designs for a remote control. How many are genuinely different from the original? Did you get stuck?
- Design cognition studies designers’ thinking when creating ideas and solving problems
- It focuses on the factors, challenges (fixation, bias), and practices that shape this ability
Four cognitive processes in design
Good designers move between four distinct processes
- Divergent thinking: identify distant, novel solutions (find more solutions)
- Convergent thinking: refine iteratively toward a better solution
- Reconceptualizing problems: rethink goals and constraints to get a better problem
- Reorganizing the design situation: externalize thinking to set up better conditions
The four cognitive processes: diverge for more solutions, converge for a better solution, reconceptualize for a better problem, reorganize for better conditions.
Problems and solutions, together
- Designers rarely “just solve a problem”—they alternate between defining problems and generating solutions
- They “crack the problem” until problem and solution converge—the moment of making sense
- The recurring dilemma is exploration versus exploitation: when to stop diverging and start drilling down
- Shift too early and you miss superior solutions; shift too late and you cannot finish
The paradox of expertise
- Iteration is the most pervasive facet of design thinking
- Being an expert does not mean generating more ideas—experts often generate fewer
- But their ideas tend to be better, because they spend effort redefining the problem first
- They are also better at strategizing when to explore and when to exploit
Cognitive heuristics and biases
- A cognitive heuristic is a rule of thumb for a quick solution—handy, but it hides other options
- The availability heuristic: ideas that come to mind easily get undue weight
- This produces bias: undue attention to a small slice of the design space
A menagerie of biases
- Anchoring: centering the design around a known reference solution
- Decoying: a reference point warps how we see an alternative behind it
- Status quo bias: undue weight to a prevailing or popular design
- Bandwagon bias: following peers’ solution paths
- All of them help us produce ideas quickly—and all limit creativity
Design fixation
- Fixation is being mentally locked into a particular solution, unable to generate alternatives
- Stronger and more constraining than bias—an inability to release an idea
- Even expert designers, taught about fixation, still exhibit it in studies (though they notice it more)
- Breaking fixation: actively create new reference points—collect examples, seek metaphors and analogies, incubate ideas, draw from art and nature
Generating solutions
A well-defined design task has four constituents
- Design decisions: the open decisions toward a design
- Design space: all designs still under consideration
- Objectives: properties an acceptable design must have (ease of use, cost)
- Constraints: hard limitations (all commands must fit in the menu)
Why design is not just optimization
- Optimization is an idealized account of problem-solving—define everything precisely, then search systematically (Ch 39)
- But design is rarely like this
- The designer must overcome bias, fixation, and the exploration–exploitation dilemma
- Quantity drives quality: generate up to 100 low-fidelity candidates per hour, suspend criticism, then select, reflect, and refine
Quantity drives quality: 50 preliminary designs distilled to 25, then to 2 final designs over roughly 30 iterations (a cashless-society project).
Ideation methods
- The eight most common: brainstorming, function/morphological analysis, scenarios, conceptual maps, checklists, analogies, metaphors, storyboards
- All follow expansion (generate) then selection (distill)
- Divided by three considerations
- how far associations are sought (further = more novel, less often valuable)
- how other people are involved (more people \(\neq\) more good ideas)
- which representations are used (visuospatial sketching vs verbal brainstorming)
Aside: how to run a brainstorm
Kelley and Littman (2001) make recommendations, popular in HCI
- Sharpen the focus (better too clear than too vague)
- Playful rules; number your ideas (aim for 100 per hour)
- “Build and jump” on each other’s ideas
- The space remembers—use tables, walls, whiteboards
- Stretch your mental muscle; get physical
- Brainstorm-killers: letting the most senior person speak first, insisting on experience, not being playful
Sketching
- Sketching has assumed a special role in interaction design—quick, disposable, low-fidelity
- It is not merely drawing an idea already in your head—it is a creative act that produces new ideas
- It supports both divergent and convergent thinking
- Greenberg and Buxton warned that usability evaluation can fixate designers prematurely; sketching alleviates this
- Drawing a sketch forces key decisions, reducing the degrees of freedom for the rest