HCI Exam 1
Interview questions
Who, what, where, when, why, how * be aware of biasing interviewee to agree with you * make sure participant is doing vast majority of the talking * organize the interview
Needfinding in your chosen area
looking at designing for a virtual interaction; the goal is to understand how people perform tasks right now without your interface, so still observe a user in their naturalistic environment.
declarative learning
requires attention, awareness, and reflection in order to attain knowledge that can be consciously recalled (mental practice)
Design Principle: Discoverability
"Is it possible to even figure out what actions are possible and where and how to perform them?" - minimize user's memory load by making actions visible - instructions for use of the system should be visible or easily retrievable whenever appropriate - the visibility principle - design should make all needed options/materials for a given task visible without distracting the user with extraneous/redundant info Advocates that functions be visible to the user so they can discover them rather than learn them elsewhere
Need finding process
"What the user really needs" Develops a deep understanding of the task. Come in with no preconceived notions about how to solve the problem. - who are the users, where are the users, what is the context of the task, what are their goals, what do they need, what are their tasks, what are their subtasks where did data conflict after going through the data gathering process? Revisit the inventory with the newly gathered data: 1. who are the users? 2. where are the users? 3. what is the context of the task? 4. what are their goals? 5. right now, what do they need? 6. what are their tasks? 7. what are the subtasks? Process of Needfinding: Conduct needfinding activities e.g. surveys, interviews, naturalistic observation -> break down the data to represent the needs -> define the requirements
Tips for avoiding need finding bias
* Confirmation Bias - "to see what you want to see; preconceived beliefs". Look for signs that you are wrong and involve multiple individuals. * Observer Bias - Separate experimenters from the motives of the participants, have others audit your interview scripts * Social desirability Bias - Conduct more naturalistic observations and record more objective data * Voluntary response Bias - Oversampling more extreme views * Recall Bias - People aren't good at recalling what they said or what they felt from past events. Utilize multiple forms of Needfinding
Design Principle: Simplicity
* Dialogues should not contain info that is irrelevant or rarely needed * extra units of info compete with the relevant units of information and diminished their relative visibility * the design should make simple, common tasks easy, communicating clearly and simply in the user's own language and providing good shortcuts
Direct Manipulation : Direct Engagement
* Direct Engagement - an example would be the moving files metaphor ; moving files should be physically moving a representation of the files. If playing a game we should be * directly * controlling our character. If we are navigating channels we should be physically specifically selecting clear representations of the channels that we want. "The systems that best exemplify direct manipulation all give the qualitative feeling that one is directly engaged with control of the objects - not the programs, not with the computer, but with the semantic objects of our goals and intentions.
Direct Manipulation : Distance
* Distance - distance between user's goals and the system itself (ideas of the gulfs in the context of feedback cycles) "the feeling of directness is inversely proportional to the amount of cognitive effort it takes to manipulate and evaluate a system." - semantic distance: difference between user's goals and their expression in the system - articulatory distance: distance between said expression and its execution, how hard it is to do what you KNOW how to do e.g. move your finger goals -> intentions -> action specification -> execution -> inter-referential I/O -> perception -> interpretation -> evaluation.
Participant observation
* Don't over-represent your participant experience, leverage your participation to inform future user participant observations Look for participant's hacks and workarounds to a task they are performing - Ask why things happen the way that they do? - recruit people to participate in a study during the task they are performing or target a relevant area
Rules for group brainstorming
* Expressiveness - share even the weirdest ideas out loud * Non-evaluation - don't criticize any ideas * Quantity - brainstorm as many as possible * Building - try to build on others ideas - stay focused - no explaining ideas - revisit the problem often - Encourage others
Design Principle: Feedback
* Feedback must be immediate and informative * Error messages should be expressed in plain language * design should keep users informed of actions or interpretations, state or condition, and errors and exceptions
Defining the requirements
* Final step for need finding is to define the requirements; specific and evaluable Requirement examples: 1. functionality 2. usability 3. learnability 4. accessibility of the interface. with other requirements generated by external project requirements 1. compatibility 2. compliance (user privacy) 3. cost
Brainstorming
* Generate an expanse of short, high level ideas to solve the problem * Research indicates to start with individual brainstorming (groups tend to coalesce around ideas early, brainstorming is most effective when generating a large amount of ideas) * Some of the best ideas start with stupid ideas
Cognitive load
* Important to know what else is competing for the cognitive resources that users are experiencing when using an interface to design a better one. 5 tips for reducing cognitive load * Use multiple modalities: When only one system is engaged it can be become overloaded while any others become bored; describe things verbally and present them visually * Let the modalities complement each other: Focus on letting each modality support, illustrate or explain each other * Give the user control of the pace: Often times interfaces have built in timers on selections needing to be made which can induce stress * Emphasize essential content and minimize clutter: Principle of discoverability states that we want the user to be able to find functions available to them but that could raise cognitive load, so design interfaces that emphasizes the most common actions * Offload tasks: ask if what a user does when interacting with an interface can be added to the interface to ease the user
Design Principle: Consistency
* Lessons learned with one system transfer readily to others * new ways of doing things is only slightly better than the old * users should have to wonder whether different actions mean the same thing * general idea is to be consistent from both within and outside the interface to minimize the amount of learning that user needs to do with our interface
Major distinctions in quantitative data
* Nominal - categorical, arises from different categories. Single and multi nominal; multi being multi checkbox. Binary nominal is just one or the other option. * Ordinal - explicit ordering to categories provided. Binary ordinal e.g. Fail Pass -> explicit ordering * Interval - intervals between numbers e.g. temperature * Ratio - has a zero point Interval and ratio can also be discrete - e.g. a countable number of minutes
Design Alternatives lifecycle
* Number of choices to make and expansiveness is larger than ever before. * How to generate ideas for design * how to explore ideas further - Settling on a design idea to early is bad - Working on improving an existing interface * "if you settled on tweaking the existing design of a thermostat, you would never develop a nest" * "If you are working on improving an existing interface, try to actually distance yourself from the existing solutions, at least during the brainstorming session * flesh out ideas for interfaces you ultimately wont end up using because by doing so, you continue to learn more about the problem.
Naturalistic Observation
* Observing people in their natural contexts * Watching engagement in an activity acquires lots of information of a design. - Limited in not knowing what the users are thinking - Informative process in understanding the problem space and preparing questions to ask for later on Tips: * Take notes, targeted information and observation * Start specific, small interactions people are doing, then abstract later * spread out sessions, observe in shorter sessions 10-15 minutes and growing past exercise reflection informs later sessions * find a partner * look for questions to investigate further Analyzes data being produced already on its own and observe it and capture it while its taking place No need for human participants when we are doing research
problem space vs design space
* Problem Space - defining a problem as far as possible away from current existing interfaces(that attempt to address the problem) when performing a task * Design Space - area in which we design solutions to the problem, as we design, the space of possible ideas might expand e.g. voice interfaces, wearables. Goal should be to explore the possible design space and grow the space and generate an expanse of ideas and designs.
User modeling
* Start to look at how the goals, operators, methods, and selections of rules map up to our ideas of design alternatives; how easy are the goals to achieve between the design alternatives? * Personas - personal and meant to give empathetic view of user's experience, user models are objective and meant to give a measurable and comparable view of user's experience Advantages: - targets general workflows - captures activity over time
Design Principle: Perceptibility
* System should always keep users informed about what is going on through appropriate feedback within reasonable time
User Types
* Understand the full range of users for whom you are designing for Identify different types of users
Design Principle: Tolerance
* Users shouldn't be at risk of causing too much trouble accidentally * Design should be flexible and tolerant by allowing undoing and redoing while also preventing errors wherever possible * design minimizes hazards and the adverse consequences of accidental or unintended actions
The problem space
* Where is the task occurring * What is going on * What are the user's explicit and implicit needs - Understand the scope of the space we are looking at. - Take into consideration a broader view of the problem space, don't focus too narrowly on the interactions the user is having with the particular interface.
Mental models
* an internal, simulatable understanding of external reality * a good interface gives a user a good mental model of the system it represents * good representations can help user's achieve strong mental models * generate expectations or predictions about the natural world * mental simulation from the mental model of the world helps to make predictions * a users mental model of a system should be an accurate depiction of how the interface is supposed to work 2 ways: predictable and learnable systems
User Error
* any user error is a failure of the interface to perform the proper action * User Slip - The user has the right mental model but does the wrong thing anyway * User Mistake - The user has the wrong mental model and does the wrong thing as a result Types of Slips: Action based slips: performed - user performs wrong action or right action on wrong object Memory lapse slips: user forgets something they knew they had to do Types of Mistakes: Rule based mistakes: user correctly assesses state of the world, but makes wrong decision based on it Knowledge based mistakes: user incorrectly assesses state of world in the first place Memory lapse mistakes: forgetting to fully execute a plan
Design Principle: Ease and Comfort
* appropriate size/space is provided for approach/reach/manipulation and use regardless of user's body size, posture and mobility.
Design Principle: Structure
* design should organize the UI purposefully in useful ways based on clear, consistent models that apparent and recognizable to users Organize our UIs in ways that help the user's mental modal map the actual content of the task - borrows from design elements from pre-computing era - the ideas used in developing magazines/newspapers are still used in developing websites
learning curves
* learning curves - "rapid learning curve" expertise grows very quickly in very little time developing experience * consistent with existing conventions (used across multiple apps) and use analogies that users understand
Design Principle: Constraints
* prevent the user from performing erroneously in the first place * designing so that users are immediately comfortable in novel situations is a goal of good UI design 4 types of constraints: * physical * cultural * semantic - inherent to the meaning of a situation * logical - self evident based on a situation
Types of evaluation
* qualitative - what they like/don't like * empirical - Perform controlled experiments, need lots of participants and address qualitative feedback first * predictive - evaluation without users, user eval is slow and expensive, doing evaluation from a day to day basis
Design Principle: Mapping
* relationship between the interface and the world * e.g. cut, copy, paste - natural mapping between our vocabulary and the system * strong mappings help make info appear in natural and logical orders mappings vs affordances - affordance is about creating interfaces where their designs suggest how they should be used. Mapping refers to creating interfaces where their design makes it clear what the effect will be from using them
Design Principle: Affordance
* relationship between the properties of an object and the capabilities of the agent that determine just how the object could be possibly used * the presence of an affordance is jointly determine by the qualities of the object and the abilities of the agent that is interacting
Design Principle: Equity
* the design must be useful and marketable to people with diverse abilities * avoid segregating or stigmatizing any users Equity - about helping all users have a good UX Flexibility - about being a means to achieving Equity E.g. make the interface BOTH : - 'discoverable' for novice users -'efficient' for expert users
learned helplessness
* the tendency to fail to act to escape from a situation because of a history of repeated failures in the past * user learns of the belief that there is no mapping between their input and the output they receive
Qualitative vs quantitative data
- Quantitative: Data supports formal tests, comparisons, and conclusions - Qualitative: Descriptions, accounts, observations - flexibility and a broader + more general picture but harder to generate more formal conclusions (prone to biases)
Stakeholders
- primary stakeholders directly interact with a tool. - secondary stakeholders receive the output of a tool ; interact with the output of the system - tertiary stakeholders are affected by the existence of the system - user centered design is also about the impact of the design on not just the end-user but also all of the stakeholders
HCI Design life cycle
-need finding -design alternatives -prototyping -evaluation cycle back to the beginning
Tips on evaluation
1. Efficiency - evaluate how long it takes users to accomplish a task; predictive models or time users in completing the task 2. Accuracy - how many errors to users commit while accomplishing a task 3. Learnability - Sit a user down in front of a UI, how long does it take a user to meet some level of expertise with the interface. 4. Memorability - refers how how much a user remembers over time. 5. Satisfaction - User's enjoyment of the system or the cognitive load they experience while using the system. Clearly articulate at the beginning - what you are evaluating - what data you are gathering - what analysis will you use Should match up to address research questions.
5 tips for group brainstorming
1. Go through every individual idea - bring ideas to group session and go through each other 2. find the optimal size - no more than 5 people 3. Set clear rules for communication - no one should block others ideas 4. set clear expectations - set session time limit/ dictate certain amount of ideas to generate 5. end with ideas not decisions - let ideas percolate before coming back to choose ideas to pursue as decisions
5 tips - mental models for learnable interfaces
1. Predictability 2. Synthesizability - see sequence of actions that lead to current state 3. Familiarity - affordances; interface should leverage actions from real-world experience 4. Generalizability - knowledge of one UI should generalize to others 5. Consistency - similar tasks/operations within a single interface should behave in the same way * helps user leverage existing interfaces of other designs
Tips of effective individual brainstorming
1. Write down the core problem - keep this visible; helps to remain grounded and focused while being creative 2. Constrain yourself - Decide on one idea that you want in a number of different categories, try to think of at least 3 ideas with non-traditional interaction methods i.e. touch and voice 3. Aim for 20 - Dont stop until you have 20 ideas, could be simple one sentence descriptions of what you want to do 4. Take a break - leave and come back, stop brainstorming and continue a couple of days later 5. Divide and conquer - dealing with bigger problem should be divided into smaller problems. Brainstorm solutions to each individual smaller problem.
Methods for design lifecycle
1. interviews, surveys, naturalistic observation 2. brainstorming, personas, storyboards 3. paper prototypes, wireframing, wizard of Oz 4. Qualitative, empirical, predictive
Challenges in group brainstorming
4 group behaviors that can block progress- 1. Social loafing - people often don't tend to work as hard in groups as individually 2. Conformity - people in groups tend to want to agree, participants are forced down the same line of thinking and can prevent convergent thinking 3. Production Blocking - Some individuals can dominate the conversation and cause others to be less heard. 4. Performance matching - people tend to converge in terms of passion and performance which can lead to a loss of momentum over time. Can often sap the energy of those who enter with enthusiasm. 5. Power dynamics - a passive pressure can build on suggestions and dampen brainstorming when done with those with more authority than you.
Scenarios
Advantages of scenarios: * includes the task context * captures activity over time * attempts to capture potential edge cases
Representing the need
After representing the user's needs, formalize into a design. Create steps of user engaging into the task via a diagram or a hierarchical network. Could further detail by breaking down the structural relationships of the components in the system and how they interact/ how they give feedback back to the user. Data gathered here can be used to summarize a more formal task analysis
Surveys
Allow for a much larger number of responses more quickly and questions can be phrased objectively Not as thorough as interviews but allows for acquiring large data sets cheaply Designing surveys: - less is more - be aware of bias : dont pressure the answer to the question - tie them to the inventory : ensure survey questions connect to some data that you want to gather - test it out : have coworkers or colleagues test out your survey - Iterate : survey design is like interface design, see what works and revise accordingly - be clear; instead of 1 to 5 say 'highly dissatisfied' to 'highly satisfied' - be concise; ask questions in plain language - be specific; ask a series of smaller more concise questions, don't ask double-barrel questions(asking two questions at once) - be expressive; emphasize the user's opinions - closely resemble the complexity of user's opinions - be unbiased and be usable; leave open-ended questions open
Design life cycle as feedback cycles
Brainstorming and designing interfaces produces outputs that are then fed back in to the lifecycle. The interface is the tools to build and evaluate interfaces, and our goal is to help a user accomplish their goal.
Apprenticeship
Capture what is occurring during the task's context because you get the user's thoughts while engaging with the task and afterwards.
errors/hacks
During naturalistic observation or with recruitment, look for when the user's *mental model* is 'weak' with an interface.
Evaluation timeline
Evaluations start as formative; primary purpose is to redesign and improve an interface Summative is later stage evaluation - intention is to conclusively say at the end what the difference was with the interface Earlier evaluations are interpretive and informal; qualitative. predictive informs how to improve interfaces over time
Needfinding
Gather a comprehensive understanding of a task the user is trying to perform who, what, where, when, why
Invisible interfaces
Interfaces don't become invisible just from good design, but also because users learn to use them. Many users become sufficiently comfortable with an interface to the point of feeling invisibly integrated with the task. 5 tips for designing invisible interfaces: 1. Use affordances - places where the visual design of the interface is how it should be used e.g. buttons are for pressing, dials are for turning 2. Know your user - invisibility means different things to different people. Beginner: interactions feel natural Expert: interactions maximize efficiency 3. Differentiate your user - provide multiple ways of accomplishing tasks e.g. right click for copy paste also hotkeys ctlr C and ctlr V 4. Let your interface teach - When users select copy and paste from the right click menu they see the hotkeys and thus are taught a better way of performing the action 5. Talk to the user - If the user is talking about the interface while the subject should be the task, then that means the interface is probably visible and means there is interference with accomplishing the task.
procedural learning
Learning ways of doing things rather than learning about specific events. Procedural learning is typically not governed by conscious controlled processes. "Unconsciously competent" due to difficultly conveying subconscious procedural knowledge as conscious declarative knowledge
Personas
Personas - represent different elements of our design/task that represent the range of real people that we care about, includes the user's context timeline - take a persona and stretch it out over time e.g. Advantages of User Profiles - includes the user's context and delineates the target audience Timelines: Captures activity over time device to exercise -> go to park -> set up audiobook -> exercise -> turn off audiobook - What prompts persona to engage in task in first place? - What actions lead up to the task? - how are they feeling during every stage of the task and can we use that? - how would design alternative impact experience throughout process? (see section 3.4 - Timelines) storyboards - sequences of diagrams of drawings that outline what happens during a certain scenario (a particular person interacting in a particular way with particular events vs routine interaction which is the purpose of timelines)
Evaluation design
Series of steps to perform to ensure useful evaluation 1. Define the task 2. Define performance measures - how will performance be evaluated? Ask users questions about their experience; avoids confirmation bias - forces us to look at it objectively Qualitative and Quantitative approaches help 3. Develop the experiment - How will we find user's performance during performance measures. Are our assessment measures reliable and valid? Qualitative and Quantitative approaches help 4. Recruit Participants 5. Do the Experiment 6. Analyze the Data 7. Summarize the Data
Direct Manipulation
The user should interact directly with their task e.g. touch screens ; a heuristic that measures how to manipulate their task as closely as possible * essentially narrows the gulf of evaluation and gulf of execution the most that it can * leveraging real world experiences (metaphors) and expectations narrow the gulf of execution also invisible interface - an interface that disappears between the user and their task, the user spends no time thinking about it, and instead spends that time thinking about the task rather than the interface VR is an example of direct manipulation, however only visual and auditory at most, there is still a lack of feedback kinesthetically.
Qualitative data
Transcripts, Field notes, Artifacts More expensive to analyze and more prone to biases Qualitative Data -> coding -> Quantitative Data
Iterative needfinding
Use some methods of need finding to inform other methods. Ask a question, interview people for a response, which will then motivate a survey which can inform the next round of naturalistic observation. In the same way as the higher-level design lifecycle, creating prototypes and evaluating them might serve as output to cycle back into need finding
Post-event protocol
Wait to get the user's thoughts until immediately after the activity such that their think aloud does not deliberately change the way they act during or before the activity.
Design Principle: Flexibility
the design should accommodate a wide range of individual preferences and abilities