HCI Final prep pt 2

ohsobeezee23's version from 2018-04-01 18:44

Summary of layout

Question Answer
What does good layout involveboth design and computational concerns
Design concern 1Leveraging properties if perception to create structure
Design concern 2Considering content and intended meaning when making grouping and alignment decisions
Design concern 3Creating hierarchies to facilitate scanning
Design concern 4Using grids to help organize screens and promote consistency
Computational ConcernsAlgorithms needed to maintain optimal visual layout across multiple window sizes, Also Algorithms differ in their complexity and flexibility

MVC (Model-View-Controller)

Question Answer
How do we design interactive software applications from a code perspectiveMost common architecture MVC YO
Primary issue with designing a spreadsheetdata and its presentation are tightly coupled, dependencies exist between underlying data and its presentation. Makes it more difficult to swap data sources, or visual rep of the data sources.
Components of mvcmodel, view, controller
Modelmanages the data and its manipulations
ViewManages the presentation of the data
ControllerHandles user input
How does MVC work in theoryView obtains pointer to model, view querrie model to construct its presentation, view registers to be informed of model changes, Controller interprets user actions to manipulate the model, model fires event, to notify interested parties of changes, View updates to match new state of model
MVC in practiceController and view tightly coupled because user ations are difficult to interpret without details of the presentation, Actions must be interpreted in terms of presentation, View has the info to map points in view to application objects
Separate concerns enables1. Multiple, simultaneous views of data

2. Alternative forms of interaction with same underlying data

3. Greater reuse of components (e.g., same component, different model back-ends)

4. Easier testing of underlying data (model) manipulations
What does seperate model make easierfor other components to make use of the data and Reduces dependencies on GUI code that could change
Separate model advantages enables application writers to more easily fit their unique app data to stock controls, and it encourages re-usability
MVC summaryPromotes flexibility and code reuse, Promotes scalability, Helps with testing


Question Answer
Main goals of evaluation1. Assess extent and accessibility of system’s functionality
2. Assess the user’s experience with the interaction
3. Identify any specific problems with the system
Assessing Functionality questionsDoes the interface provide functionality users require?
Is functionality clearly reachable?
Does it match their expectations about how the task should be performed?
Two main types of evaluationformative evaluation and summative evaluation
Formative evaluationdone at different stages of development to check that the product meets users’ needs
Summative evaluationassesses the quality of a finished product

Evaluation approaches

Question Answer
5 Eval approachesQuick and dirty, Analytical, Usability testing, experiments, field studies
Quick and dirtyCommon practice for getting informal feedback, feedback is either from knowledgeable consultant or small number of users
Emphasis for quick and dirtyget input quickly rather than formal eval methods with carefully documented findings
Analytical evaluationEval without involving end users.
Two types of analytical evalInspection (Experts asses the system) and performance modeling (Theoretical models are used to predict user performance)
Usability testingUsers are observed using the system in a controlled environment
Typically asked to perform a set of prescribed tasks
Performance measured (e.g., speed, accuracy, etc)
Interviews/questionnaires administered to assess subjective impressions
Field StudiesUsers use the system in its natural setting
Most data gathered through observation and interviews
Systems can also be instrumented to collect usage data

Decide Framework

Question Answer
What does decide stand forDetermine, explore, choose, identify, decide, evaluate
Determinethe goals of evaluation addresses
Explorethe specific questions to be answered
Choosethe eval method and technique to answer the question
Identifythe practical issues
Decidehow to deal with the ethical issues
Evaluateinterpret and present the data

Analytical Evaluations

Question Answer
2 types of analytical evaluationsinspection (heuristic and cognitive walk-through) and performance(Fitts’s Law, Hick-Hyman Law, Keystroke Level Modelling, GOMS)
Inspection also known as asking expertsask a bunch of experts if users are too expensive to critically analyse the product, Or heuristic review
Cognitive walk-throughsimulate users problem solving process at each step. Check if user goes to next correct action
3 questions to ask when doing cognitive walk-throughWill the user know correct action, Will the use notice how to do the correct action, Will the user associate and interpret the response from the action correctly

Heuristic evaluation

Question Answer
What is Heuristic evaluationSystematic inspection of an interface design to see if an interface complies with a set of usability heuristics, or usability guidelines
General process3-5 inspectors usually experts, inspect interface in isolation, and results are aggregated afterwards
Advantages1.Fewer practical and ethical issues to deal with
2.Best experts have knowledge ofapplication domain & users
3.Can be difficult & expensive to find experts
4.Could have disagreement between experts
Problems with Heuristic evalprinciples may be too general, designer might be too petty, experts disagree, can be hard to find experts, NOT A REPLACEMT FOR USER TESTING
Performance modeling mathematical models to generate quantitative predictions of certain interface actions or sequences of actions
Fits lawpredicts how long it will take users to acquire targets once they know which target to select
Hick-Hyman lawmodels the time it takes users to decide between n alternatives
Hick-Hyman law distinctionModels decision time, not searching time. If the user is not familiar with the interface elements: time to search through n items is linear not logarithmic
Keystroke Level ModelModels performance given a sequence of steps for an expert user
GOMSGoals, operators, Methods, Selection rules
Describe GOMShigher level than KLM, the input is a detailed description of UI and tasks, and output is various qualitative and quantitative measures
Performance modeling AdvantagesCan evaluate components of interface prior to building it
Good for comparing different interface possibilities
Evaluations with human subjects are expensive and experts can be hard to find
Performance modeling ProblemsDifficult to model complex tasks, Most models consider only expert behaviour, For really accurate predictions, coefficients need to be determined empirically

Nielsen's heuristics

Question Answer
Visibility of system statusAre users kept informed at all times?
Match between system and real worldIs the UI language simple?
User control and freedomAre there easy escapes from unexpected locations?
Consistency and standardsIs performing similar action consistent?
Help users recognize, diagnose, recover from errorsAre error messages helpful?
Error preventionIs it easy to make errors?
Recognition rather than recallAre objects, actions and options always visible?
Flexibility and efficiency of useAre there accelerators?
Aesthetic and minimalist designIs any unnecessary and irrelevant information provided?
Help and documentationIs help provided that can be easily searched?

Usability Testing

Question Answer
What is usability testinga formal method to evaluating whether a design is learnable, efficient, memorable, can reduce errors, meets users' expectation, etc
Key thing to remember about usability testthe design is being evaluated not the users
Rough outline for usability test1. bring in real users 2. Have them complete tasks with your design, while you watch 3. Measure and record things such as task completion time, errors, satisfaction, problems etc.
What to do with data collecteduse data to Identify problems, provide design suggestions to design/ engineering team, iterate on design, repeat
Important considerations for usability testingusually takes place in a usability lab or other controlled space, Major emphasis is on selecting representative users, and developing representative tasks, 5-10 users typically selected, should take no longer than 30 minutes
Pilot StudiesEspecially important for usability testing

Make sure your plan is viable

All the corners are checked (your script,questionnaires, tasks, etc., all work)

It is worth doing several to iron out problems before doing the main study

Ask colleagues if you can’t spare real users
Creating a tasktask is defined to probe a problem and should be straightforward
How many participants are enough for usability testingusually 5-10

Usability testing summed up

Question Answer
Usability testing is a form ofapplied experimentation
users are brought into controlled environments tocomplete focused tasks
Focus of the testing ison having representative users and tasks
Typically both ....objective and subjective data is collected
Important end goaldraw meaningful conclusions about your system's current strengths and limitations
Identify practical issuesplan to run tests, Have scripts in place, Test equipment,Have recording material prepared
How to deal with ethical issuesConsent forms
Elements of usability testingEvaluate, analyze, and present data
How to present the resultsrank in severity, small suggestions on hot to address each problem, provide evidence

Data analysis (ties in with usability testing)

Question Answer
Qualitative dataCollected from interviews, some types of questionnaires, observation notes
Interpreted & used for telling a ‘story’ about what was observed
Quantitative dataCollected from interaction & video logs
Presented as values, tables, charts, graphs and treated statistically
When to use Qualitativeused when we don’t know what to expect, to define the problem or develop an approach to the problem.
When to use Quantitativewhen we already know the problem, and want to project results on a larger population
How to make sense of your dataaffinity diagrams, discussion with others who worked with you
Likert scalesused for measuring opinions attitudes, beliefs
How to design Likert scalescreate a pool of short statements and then make some logical/conceptual groups and decide on a scale
When choosing scale numberseven for when you need them to lead one way or another odd if you want a neutral. Also if majority is uninformed about topic use small number, if experts use lager set
Anchorsverbal comments above the numbers
Analyzing questionnaire datapresent results clearly, and simple stats mean a lot
Observing peoplemake sure you use the think aloud method, or to remove awkwardness do the co-discovery learning where a semi-knowledgeable coach sits there as well
When observingbe quiet watch and understand
SummaryUsability testing is a form of applied experimentation Users are brought into controlled environments to complete focused tasks Focus of the testing is on having representative users and tasks Typically both objective (e.g., task completion times,errors) and subjective data is collected (e.g., questionnaire data) Important end goal: draw meaningful conclusions about your system’s current strengths and limitations


Question Answer
Goal for usability testingCheck that the system being developed is usable by the intended user population for their tasks
Experiments goalTest hypotheses to discover new knowledge by investigating the relationship between two or more things
This is pretty much the scientific method so no more pointers just knowindependent variables, dependent ,hypothesis, relationships, nuisance variables.
Types of validityInternal, External, Ecological
Internal validitythe changes in dependent variables are caused by the independent variables
External validityResults can be generalized to other settings, populations, tasks, etc
What is the trade off between internal and external validityThe more tightly you control the experiment, the less generalizable the results
Ecological validityTo what extent do the study conditions mimic those in the real world. Related to external validity, but not the same
2 Study design decisionsbetween-subjects and Within subjects
Between Subjectsalso known as randomized experiments each participant does one of the experimental conditions, no learning effects, doesn't account for individual variability
Within SubjectsEach participant completes all experimental conditions, better able to account for individual differences, requires fewer participants, allows participants to make direct comparative statements, learning effects are possible.
The more participantsthe better.

Universal Design

Question Answer
What universal design means essentiallyMake sure everyone can use your technology account for all sorts of disabilities, NO SUCH THING AS THE AVG USER
Age related disabilities (might come up on final) Decreased visual acuity, Problems with low levels of illumination, Reduced ability to distinguish colors, especially in short wave lengths , Visual search is harder, takes longer etc
Why design for users with disability ?It's ethical, the law, business opportunities and because the curb cut phenomenon
Curb cut phenomenonSidewalks needed to be cut to accommodate users with disability, ended up benefiting everyone, strollers, luggage, bikes etc
Economic exclusion People are excluded if they cannot afford some essential technology.
Cultural exclusionAssumptions embedded in technology that exclude certain cultures.
Social ExclusionWhen tech is un-available at an appropriate time and place or if people are not member of particular social group and cannot understand particular social mores or msgs
Physical exclusion Input and output devices making excessive demands on user abilities (ex. ATM may be positioned too high for a person in a wheelchair to reach)
Conceptual exclusionPeople may be excluded because they cannot understand complicated instructions or obscure commands or they cannot form a clear mental model of the system.

Recent badges