Interaction Design

Interaction design cycle

1) establish requirements 2) develop alternatives 3) prototype 4) evaluate 5) iterate

Core characteristics of interaction design

1) users should be involved in the design process 2) specific usability and user experience goals should be identified, agreed upon and documented at the start of the project

Low fidelity prototype

Use materials unlike what you would use for the final mockup Cheaper and easier to change

High fidelity prototype

Use materials you'd expect in the final product. Can use actual software and hardware

Theoretical frameworks for qualitative design

Active theory Distributive cognition

Field studies

Aim to understand what users do naturally in their environment Used to identify opportunities for new technology and how to implement them

Grounded theory

Aims to create a theory from the systematic analysis of data . Based on categorisation approach called coding

Advantages and disadvantages of using windows

Allows you to view lots of data at the same time and multiple tasks can be performed together Hard to switch between windows without getting the user distracted Multiple windows makes it hard to focus and find the right one

Usability testing

Applied to experimentation phase to check if system is usable by audience

Stages of heuristic evaluation

Briefing to inform experts what to do Evaluation period - experts work separately, take a pass to get a feel for the product and then take a second pass to look more in depth into the product Debrief with other experts to prioritise the problems

Personas

Capture a set of user characteristics. Should not be idealised

Thematic analysis process

Category construction Sorting, expanding and reducing categories Check back again with data Compare with co workers

Purpose of data gathering

Collect data that is accurate and sufficient enough to produce set of requirements. You can use data to also evaluate product based on feedback from users

Interface types

Commands Speech WIMP and GUI Data entry Multimedia Etc

Direct manipulation

Continuous display of objects and actions of interest. User uses the application solely through clicking buttons etc

Types of evaluation

Controlled settings with users - usability testing Natural settings with users - get feedback of product from users in natural user environment Settings without users - predict and analyse interface analytics

Goals for interaction design

Develop products that are easy to use and learn. It's important to involve users in the process

Cognitive prosthetic devices

Devices that reduce the need for the user to remember by allowing you to search up the stuff e,g. Google

Reading and speaking

Different people prefer reading/ listening/ speaking more or less than others Design implications Speech based menu options need to be quick and simple Allow larger text modes Add accents to artificial speech to make it easier to understand

Unstructured interview

Don't follow any script Can't be easily replicated Very rich but may go off topic

Advantages and disadvantages of online questionnaires

Easy to distribute Can get a lot of responses easily Low costs Errors can be corrected easily People can change their answers No easy way to prevent users from submitting more that one questionnaire Hard to sample data if population is unknown

Heuristic evaluation

Evaluate user interface against recognised usability principles ( heuristics)

Inspection methods

Experts use their knowledge and experience of users and the field to review usability On average you only need 5 experts to get 75% to 80 % of errors Heuristic evaluation Cognitive walkthroughs

Active theory

Explain human behaviour based on what we do in the world Helps identify tension between two objects Two main models - - one that outlines meanings of activities - one mediates role of artefacts

Theory

Explanation of phenomena. Can help with identify factors

Interviews

Face to face usually Can be open ended or close ended Good because it encourages contact between developer and audience Bad because audience may get intimidated by intimate setting

Wizard of Oz prototyping

Facilitator: provides product, creates tasks and gives instructions to participant. Notes down how participant interacts with product Participant: operates the user interface Easier to operate than fully functional prototype User centred development Feedback delivered instantly Easy to commit to designs that are not technologically possible Can not simulate every feature

Traits of usability testing

Few users used Conditions are as controlled as possible: controlled lab space, select representative users, develop representative tasks, test conditions the same for all users Usually not replicable Results inform future design choices of product Focus on time taken for user to complete tasks and the number of errors

Structured interview

Follow script carefully Can be replicated May lack richness

Semi structured

Follow script for the most part but can expand on interesting questions

Vision

Goals that drive research and development

Direct observation in controlled environment

Good to get understanding about what users do Results might not be representative of real life

Direct observations in the field

Good to understand context of user activity Ethnography: immerse yourself into a new culture without any previous judgements or prejudice Ask questions like what, where , how, why Time consuming to collect data

Things to consider when evaluating data

Reliability - is data replicable Validity - does data do what we need it to Ecological validity - does environment where tests are done effect the results Bias - are there biases that'll distort the values Scope - how generalisable are the results

user experience

How a product behaves in real life and how people respond to it. You can't create a user experience, you can only design for one

Memory

How information is encoded into our brains and retrieved from our knowledge Design implications Don't overload users memory with complicated instructions Encourage recognition over recall Use ways to help encode information

Perception

How information is gathered and turned into experiences Design implications Make text legible Icons are easily distinguishable Spacing used effectively to show different sections

User experience

How the product makes the user feel. Helps designer know the nature of the product based on the emotions they want the user to emote Desirable traits- satisfying, easy to use, fun, entertaining Undesirable- boring, hard to use, frustrating, annoying

Learning

How you learn through a computer based application Design implications Interface should have help options Should encourage exploration

How to analyse problem space

Identify problem with user experience and product Propose a better solution Evaluate new designs effectiveness

Paradigm

Inspiration for conceptual model

Interaction types

Instructing Conversing Manipulating

Conversing

Interacting with system like you're having a conversation with it Easy to learn for technophobes because they're using a medium that's comfortable to them Problem with interpreting speech accurately

Structure of interviews

Introduce yourself and the tasks you're about to go through today Warm up with some easy questions The main body where you ask the bulk of the questions in a good order Cool off period with easy questions Thank them for coming and end interview

Why take into account interaction design

Let's you understand what people want and need Identify incorrect assumptions Be aware of the strengths and capabilities of your audience so the product is tailored to them

Models

Makes it easier for designers to predict and evaluate alternative designs

Experiments for research

Many participants Replicable Strongly controlled conditions Scientific report for scientific community produced

Components of conceptual model

Metaphors and analogies: under stand how a product is used and how it's used Concepts users are exposed to: the product task domain, specifying the specific operations the user will be able to perform like saving file, loading a file etc Relationship and mappings between different concepts

Prototype

Mockup of a product that is easy to alter based on feed back from users, team and clients. Should aim to address one aspect of product functionality

Exploring

Moving through a physical or virtual space. Physical through sensors etc and virtual through VR etc

Puristic walkthrough

Multiple experts work separately through the walk through and then a carefully managed discussion takes place which leads to agreed decisions

Analysing quantitive data

Numerical data Simple analysis: numerical methods to ascertain size and magnitude

Questionaire

On paper / online

Coding

Open: identify categories and distinct themes Axial: find links between categories and sub categories Selective: form theoretical scheme

External consistency

Operations between applications etc behave the same for the similar purposes

internal consistency

Operations within the application behave the same for similar purposes

Benefits of conceptualising

Orientation: let's team question how the conceptual model will understood by the end user Open mind: prevents the team from becoming too close minded too early Common ground: allows team to find terms which they all agree on and can further debate terms they disagree on

Matched participants

Participants are paired based on traits such as intelligence etc Reduce effect of individual traits Can not take into account all individual differences

Manipulating

Physical actions like using cursor to click a button to perform action Users can quickly see the results from their action Can be useful to learn basic functionality More complex concepts can't be explained through objects and clicking objects Takes up lots of space Slower than using key board short cuts

HCI cycle

Plan user centred process Understand and specify how product will be used Specify what the user needs from the product Produce solution Evaluate design against user requirements If it's met requirements then deliver

Usability goals

Practical goals on how the application should behave Efficient - minimises user effort Safe - can recover from errors safely without data loss Utility - does what you need it to Easy to learn Memorable- easy to remember how to use

Predictive models

Provides a way to evaluate products without the need of users Less expensive than user testing Usefulness limited to tasks that are predictable with predictable input such as button presses Based on export error free behaviour

Cognition

Provides information about what a user can/ can't physically do and how to design their product optimally to Minimise effort and strain of user

Same participants

Same participant used in all different versions of test Less users needed Order is crucial as can add bias

Attention

Selecting a stimulus for your mind to focus on out of a set of stimulus Allows user to focus on more important stuff Design implications Avoid cluttering interface Make info easy to spot Use colours to highlight important information

Framework

Set of related concepts that state what you're looking for

Key issues with data gathering

Setting goals - how will you analyse the collected data to retrieve information relevant to you Identify participants: determine audience and sampling method Relationship with participant - Keep professional + get consent Triangulation - look at data from multiple different angels and collect lots of different data sets so you can get a thorough understanding of your audience Pilot studies - you can run small studies before the official one

Consistency

Similar operations and elements used for similar tasks to make learning easier e.g. Ctrl+ C , Ctrl+V etc used to perform different actions but with similar methods

Cognitive walk though

Stepping through preplanned scenarios, spotting problems Designers present product and scenarios. Tell experts about population and use of the product etc

Encoding

Stimuli given more attention to have a higher chance of being remembered

What do you do in interaction design

Take into account users needs and the activities they want to perform to produce a product that alligns users needs with activities. Optimise interactions between user and product

Instructing

Tell system what to do using commands like command line Faster and more efficient Steep learning curve

Deep vs shallow mental models

The depth of knowledge known. Knowing how to drive a car vs knowing how a car works

Participatory design

The end user should be included in the design process

FITTs principle

The time taken to point at an object is based on the distance from the object and how big the object is Smaller the object and the further away you are, the longer it'll take to point to the object

Indirect observation

Tracking users without disturbing their environment. Can collect results automatically through logging and diary entries . Can collect large amounts of data without much effort because it's automatically collected Bad because diary entries etc means reliance on subject to enter their findings regularly and accurately

Methods of qualitative data analysis

Traditional method Grounded theory

Traditional method of qualitative data analysis

Try and explain findings by basing them around existing theories Bad because basic theories can't explain complicated human behaviour

Distributive cognition

Used to describe interactions between people Cognitive system: the interactions between people, the artificial to they use and the environment they're in Communication pathway : how they interact Propagation of representative states: how information is transformed through media and how it's mediated mentally and through technology

Mental models

Users develop a mental understanding of how a product works through using it. can make inferences using mental models to figure out how to handle new and unexpected situations Requires conscious and subconscious processes

Interface metaphors

Using interface components like send buttons etc to visualise concepts like sending emails Makes learning new system easier Helps user learn underlying model Makes product more accessible Forces users to only think in terms of metaphors Can constrain designers in how they conceptualise a concept space

Uses needs

What are people good/ bad at and what might make their life easier

What questions should you ask when understanding problem space

What do you want to create What are your assumptions: traits you take for granted but need extra research Will your product achieve what you want it to

Recall retrieval

When you don't need a prompt to remember the data

Recognition retrieval

When you need a prompt to remember data

Different participants

Where participants are separated into groups and different groups experience different conditions No order effects Many subjects required and subjects have individual differences

Evaluation

Why - check user requirements to ensure product fits user requirements so it's liked by the user What - a conceptual model, early prototype and more complete prototype Where- in natural setting or in labs When - should be doing them through design process

Analysing Qualitative data

Words etc Simple analysis: look for recurring themes and categorising data. Theme based analysis

Advantages of using multimedia

You can find the right medium to enable best learning Encourage exploration of a product

Problem solving and reasoning

You use reflective cognition to think about what to do and the pros and cons of using certain methods based on the artefacts you have Design implications Provide extra info for user to make best decision Offer quick help so decisions are fast

conceptual model

an explanation for how a system works or is organised so designers can straighten out their thinking before starting the project

Claim

state something without providing evidence or proof.