SE365 · LECTURES 9 & 10Prototyping & Construction
SE365 · LECTURES 9 & 10 · SLIDE BREAKDOWN

Prototypes answer questions.
So prototype appropriately.

A prototype can be a lump of wood, a slide show, a paper sketch or working software. What matters is which question it is being built to answer - and what compromise you accepted to build it that quickly.

What is a prototypeWhy prototypeLow fidelityHigh fidelityHorizontal vs verticalWizard of OzConceptual designConstruction

LEARNING OUTCOMES

  1. Describe prototyping and different types of prototyping activity.
  2. Produce a simple prototype from the models developed during the requirements activity.
  3. Explain conceptual design and its steps from requirements to design.
  4. Evaluate an interface metaphor against the five stated criteria.
  5. Generate storyboards from scenarios and card-based prototypes from use cases.
01 - WHAT AND WHY

A prototype is anything users can react to

In other design fields a prototype is a small-scale model - a miniature car, a miniature building, a 3D-printed part. In interaction design it can be many other things.

Why prototype?What it buys you
Evaluation and feedback are central to interaction design.A prototype is the object those activities act on.
Stakeholders can see, hold and interact with a prototype.Far easier than reacting to a document or a drawing.
Team members can communicate effectively.A shared concrete referent instead of competing mental models.
You can test out ideas for yourself.Cheap disproof before expensive commitment.
It encourages reflection.A very important aspect of design.
Prototypes answer questions.And support designers in choosing between alternatives.

What to prototype: technical issues; work flow and task design; screen layouts and information display; and difficult, controversial or critical areas. Prototyping is also described along filtering and manifestation dimensions - which aspects of the design the prototype keeps, and in what form it appears.

MEMORY HOOK

The single most useful sentence in this deck: "Prototypes answer questions, so prototype appropriately." Every choice below - fidelity, horizontal or vertical, paper or code - follows from asking which question you are trying to answer.

02 - LOW-FIDELITY PROTOTYPES

Unlike the final medium, on purpose

A low-fidelity prototype uses a medium which is unlike the final medium - paper, cardboard. It is quick, cheap and easily changed.

a. Storyboards

Often used with scenarios, bringing more detail and a chance to role play. A series of sketches showing how a user might progress through a task using the device. Used early in design.

b. Sketching

Important to low-fidelity prototyping. Don't be inhibited about drawing ability - practise simple symbols. The point is the idea, not the draughtsmanship.

c. Card-based

Index cards (3 x 5 inches), each representing one screen or part of a screen. Often used in website development, and generated from use cases.

THE GENERATION RULE

The deck states two direct mappings you should be able to reproduce: storyboards are generated from scenarios, and card-based prototypes are generated from use cases. Narrative → storyboard; step sequence → cards.

03 - HIGH-FIDELITY PROTOTYPES

Materials you expect in the final product

A high-fidelity prototype uses materials you would expect to be in the final product, and looks more like the final system than a low-fidelity version.

Wizard of Oz

The user thinks they are interacting with a computer, but a developer is responding to output rather than the system. Usually done early in design to understand users' expectations. The lecture asks what is 'wrong' with this approach - the answer is that it tests the interaction concept while proving nothing about feasibility, and it depends on a human able to respond as no real system yet can.

Compromise typeWhat it provides
HorizontalA wide range of functions, but with little detail.
VerticalA lot of detail for only a few functions.
THE COMPROMISE RULE

All prototypes involve compromises - a slow response, sketchy icons, limited functionality. And: compromises in prototypes must not be ignored. The product needs engineering. A prototype that quietly becomes the product ships every compromise with it.

MEMORY HOOK

Horizontal is wide and shallow; vertical is narrow and deep. Picture the letter: the horizontal bar stretches across everything, the vertical bar drills down through one spot.

04 - CONSTRUCTION

Taking the prototypes and creating a whole

Construction means taking the prototypes - or learning from them - and creating a whole. Quality must be attended to.

Physical computing is the hardware side of construction: building and coding prototypes using electronics. The toolkits named are Arduino, LilyPad (for fabrics), Senseboard and MaKey MaKey. All are designed for use by a wide range of people, not only engineers.

MEMORY HOOK

Construction qualities: U-R-R-M-I-P-E. The first is usability and the rest are the software-engineering qualities you already know - the point of the list is that usability sits among them, not after them.

05 - CONCEPTUAL DESIGN

From requirements to design

Conceptual design transforms user requirements and needs into a conceptual model - the first step of design.

THE DEFINITION TO QUOTE

A conceptual model is "a description of the proposed system in terms of a set of integrated ideas and concepts about what it should do, behave and look like, that will be understandable by the users in the manner intended."

  1. Understand the functionalityStep 1 of finding a metaphor.
  2. Identify potential problem areasStep 2.
  3. Generate metaphorsStep 3.
Evaluating a metaphor - five questionsWhat it probes
How much structure does it provide?A thin metaphor explains nothing.
How much is relevant to the problem?Borrowed structure that does not apply becomes misleading.
Is it easy to represent?If it cannot be drawn or built, it cannot be shown at the interface.
Will the audience understand it?A metaphor is only familiar knowledge if the audience has it.
How extensible is it?Will it still work when the product grows?

Expanding the initial conceptual model asks three groups of question. What functions will the product perform - and what will the product do versus what will the human do (task allocation)? How are the functions related - sequential or parallel, and how are they categorised, such as all actions related to privacy on a smartphone? What information is needed - what data is required to perform the task, and how is that data to be transformed by the system?

Concrete design

The many detailed aspects: colour, icons, buttons, interaction devices. It must account for user characteristics and context - accessibility and cross-cultural design - and follow cultural website guidelines. The quotation to remember: successful products "are ... bundles of social solutions. Inventors succeed in a particular culture because they understand the values, institutional arrangements, and economic notions of that culture."

06 - SCENARIOS IN CONCEPTUAL DESIGN

Expressing proposed or imagined situations

Scenarios return here, but doing a different job than in requirements.

THE TWO GENERATION PATHS

Scenario → storyboard. Use case → card-based prototype. Both appear as worked examples in the deck, and both are the kind of thing an exam asks you to demonstrate rather than define.

MEMORY HOOK

A scenario is a story, so it becomes a picture strip (storyboard). A use case is a numbered sequence, so it becomes a stack of numbered cards. The artifact keeps the shape of its source.

MISTAKES STUDENTS USUALLY MAKE

Mistakes students usually make

Each claim below is the wrong answer; the line beneath it is the correction, in the wording this course marks against.

"High fidelity is always better."
High-fidelity prototypes carry the danger that users think they have a full system, and they are slower and more expensive to change. Paper prototypes are quick, cheap and very effective at identifying problems in early design.
Swapping horizontal and vertical compromises.
Horizontal = wide range of functions with little detail. Vertical = a lot of detail for only a few functions.
Ignoring prototype compromises when building the product.
"Compromises in prototypes mustn't be ignored. Product needs engineering." A slow response or sketchy icon accepted in a prototype is a defect if it survives into the product.
Thinking Wizard of Oz means the system is partly built.
The user believes they are interacting with a computer, but a developer is producing the responses. It is done early to understand users' expectations, and proves nothing about technical feasibility.
Generating a storyboard from a use case.
The stated mappings are scenario → storyboard and use case → card-based prototype.
Listing only usability as a construction quality.
The list is usability, reliability, robustness, maintainability, integrity, portability and efficiency.
Treating conceptual design as screen design.
Conceptual design produces the integrated ideas and concepts about what the system should do, behave and look like, understandable by users. Colour, icons and buttons belong to concrete design.
CHEAT SHEET

Shortest correct answers

The night-before table: every term in this lecture with the smallest answer that still earns the mark.

ConceptShortest correct answer
PrototypeAnything from screen sketches, a storyboard, a slide show, a video, a lump of wood or a cardboard mock-up to limited-functionality software.
Why prototypePrototypes answer questions, support choosing between alternatives, enable stakeholder feedback, aid team communication and encourage reflection.
What to prototypeTechnical issues; work flow and task design; screen layouts and information display; difficult, controversial or critical areas.
Low fidelityUses a medium unlike the final one - paper, cardboard. Quick, cheap, easily changed.
Low-fi formsStoryboards, sketching, card-based prototypes.
StoryboardA series of sketches showing how a user might progress through a task; used early, often with scenarios.
High fidelityUses materials expected in the final product; danger that users think they have a full system.
Wizard of OzThe user thinks they interact with a computer, but a developer produces the responses; used early to understand expectations.
Horizontal prototypeWide range of functions with little detail.
Vertical prototypeA lot of detail for only a few functions.
Construction qualitiesUsability, reliability, robustness, maintainability, integrity, portability, efficiency.
Physical computing kitsArduino, LilyPad (fabrics), Senseboard, MaKey MaKey.
Conceptual modelA description of the proposed system as integrated ideas and concepts about what it should do, behave and look like, understandable by users.
Metaphor evaluationStructure, relevance, ease of representation, audience understanding, extensibility.
Task allocationDeciding what the product will do and what the human will do.
Concrete designColour, icons, buttons, interaction devices, plus accessibility and cross-cultural design.
APPLY IT

Exam-style application

Write your own answer first, then open the model answer. These are the longer-form questions this material generates.

You have one week to decide between three navigation concepts for a banking app. Choose a prototyping strategy and justify every choice.
The question being answered is which structure users can navigate, not whether it can be built, so prototype appropriately: use low fidelity, since it is quick, cheap and easily changed, and a medium unlike the final one keeps stakeholders discussing structure rather than colour. Build card-based prototypes - one card per screen or part of a screen - generated from the use cases, plus storyboards generated from two scenarios so users can role play the whole journey. Take the horizontal compromise: a wide range of functions with little detail, because the comparison is across the breadth of the navigation. Note the compromises explicitly - no real response times, sketchy icons - because compromises must not be ignored when the product is engineered. Do not build high fidelity here: it is slower to change and carries the danger that users think they have a full system and start commenting on polish instead of structure.
Evaluate a 'filing cabinet' metaphor for a cloud storage product against the five criteria.
Structure: strong - drawers, folders, hanging files and labels give a rich set of relationships to borrow. Relevance: partly - hierarchy and naming transfer, but a physical file exists in exactly one drawer, whereas cloud files are shared, versioned and appear in several places, so the metaphor actively misleads about the most important new behaviours. Ease of representation: high - drawers and folders are trivially drawable and already conventional. Audience understanding: falling - younger users may never have used a physical filing cabinet, so the familiar knowledge the metaphor depends on is not universal. Extensibility: poor - sharing, sync conflicts and collaborative editing have no cabinet equivalent, so the metaphor breaks exactly where the product grows. Conclusion: borrow folders and naming, but do not commit to the cabinet, or it will constrain how the team conceptualizes the problem space - one of the stated problems with metaphors.
Explain why a successful prototype is not a product, using the lecture's own vocabulary.
Because all prototypes involve compromises. For software-based prototyping that may mean a slow response, sketchy icons or limited functionality, and structurally it means either a horizontal compromise - a wide range of functions with little detail - or a vertical one - a lot of detail for only a few functions. Neither is a whole system. The lecture states the rule directly: compromises in prototypes must not be ignored, and the product needs engineering. Construction is the separate activity of taking the prototypes, or learning from them, and creating a whole in which quality is attended to - usability, reliability, robustness, maintainability, integrity, portability and efficiency. A high-fidelity prototype is especially dangerous here, because users already think they have a full system, which makes it politically hard to admit that the engineering has not been done.
POP QUIZ

Check yourself

6 questions. Every option is explained after submitting, including why the wrong ones are wrong.