L3 - Feedback cycles: In- and Output Flashcards
Why feedback?
informing the user about what has happened in order to let him know what to do next
- > computer “talks” to user
- > it’s cyclic (consider the human action cycle)
- > interaction between computer and user over interface: do something, observe result, adjust behavior if necessary (computer does something and sends result of that - over interface)
How does HCI feedback take place (over which channels)?
- visual (text, color change, …)
- auditory (sounds)
- tactile (vibration, …)
- > combinations of these
How does feedback need to be?
- prompt (immediate, timely)
- clear to understand
Between which parts of the (human) action cycle does the Gulf of Execution take place? (going from user (task goal) to computer system)
- intent to act (identify goal in terms of the computer system)
- plan action(s) (identify necessary actions to accomplish goal)
- execute action(s) (within the interface)
Between which parts of the (human) action cycle does the Gulf of Evaluation take place? (going from user (task goal) to computer system)
- perceive system state (perceive form of output)
- interpret perception
- evaluate the interpretation (evaluate feedback in terms of the task goals)
What is the Gulf of Execution?
difference between the intentions of the users and what the systems allows them to do (or how well the system supports those actions)
How can the Gulf of Execution be narrowed?
- help pages, constrains, use conventions/standards
- use affordances/real-world mapping (skeuomorphic design)
- for novice user: make goal identification and actions to get there visible and discoverable
- for expert user: reduce number of steps
- > heuristic of flexibility (for different users)
- provide feedforward; showing users what will happen if action continues
- minimize effort needed to execute each action (execution as easy and useful as possible)
What is the Gulf of Evaluation?
difficulty of assessing the state of the system and how well the artifact supports the discovery and interpretation of that state
Googled:
Disparity between the user’s perception (or discovery) of the system state, and the actual system state.
How can the Gulf of Evaluation be narrowed?
- giving feedback frequently (keeping updated about system status)
- giving immediate feedback (right after action was taken)
- balance feedback with actions (no overload or big feedback for tiny action)
- vary the feedback (type and sensory change; makes context of feedback better understandable)
- being specific
- allowing direct manipulation (feeling that action is taken)
- > use visible actions (signifiers), feedforward signals and feedback
Important to consider using the gulf of execution/evaluation concept …
it’s just a model to examining the principles -> there are different models available (may also contradict)
Actions and Feedback towards input and output - their relation?
- input devices mediate input (actions) with a computer system; software gives affordance on how to use device (guides actions)
- output devices mediate feedback from computer system (device feedback informs evaluation)
- > feedback in computer systems mich more constrained than in physical devices
what are characteristics of input?
- size and shape
- wireless or tethered (without or with cable, trade off between their pros and cons)
- degrees of freedom (describes how many different types of input it can accept, allowing for direct mapping (more action possibilities) -> more intuitive)
- relative vs. absolute (concerning how precisely one is tracked in space)
- hand-based (ex. glove, controller, …)
- body-worn
- world- grounded (ex. washing machine, treadmill, bike, …)
- > choice which one is used depends on device (object)
VR: why and how - tracking and sensing?
most VR input relies on special sensors to keep track of user’s behavior
- head tracking (detect head turning)
- full body capture (detecting surroundings)
- hand worn/held
- tracked controllers
- physical input (ex. shacking the phone)
- position/orientation (VR is real world oriented and user expects that as well (things represented as in real world))
- eye tracking
- HR, EMG, GSR etc (muscle signals, skin measures to see stress for example)
- > every part of input created has to be evaluated by the software in order to generate and appropriate response (happens inside the system)
- > it’s not the computer giving the signal, it’s the software
what are characteristics of output?
- sensory channel(s)
- resolution/range
- spatialization (sounds from different directions enriches the experience)
- head mounted (ex. headphones, AR glasses, …)
- world fixed (ex. computer screen, speaker, …)
- hand-held
- body-worn (ex. smart watch)
What is meant by passive haptics and substitutional reality?
use real world objects to convince people VR experience is real
thereby, texture is important
-> a certain point of divergence is ok (tricking the brain)
What is understood under direct manipulation?
- natural end point of good feedback cycles
- > user should feel as if they directly interact with the task (ex. zooming), even though they actually interact with the interface and all the background happenings/processes are not displayed
- > similar time natural behavior (therefore, cognitive load should be kept low)
Difference between visible and invisible interface
invisible: no thinking about interaction with system (interface invisible)
- > goal of HCI
- > less consciousness involved -> more confidence user has with interaction
- > interfaces can become invisible over usage time (repeated use) -> learning it (takes time and effort)
HCI is best if it is ____? (and why?)
-> invisible
leverage prior experience, offer responsive feedback and understand the user
How is feedback expected in VR?
- user is grounded in real-world (not expected to feel button it sense if self-motion)
- > user fully immersed in simulation; expecting sensory experience veridical (wahrheitsgetreu) to real-world
What is meant by sensory integration?
- we perceive world through multiple sensory channels
- we integrate sensory information to interpret meaning
- > let’s us understand ongoing situation
- > if they come into conflict it results in discomfort or cyber sickness
- > VR creates discomfort if feedback conflicts with user expectation
How to minimize sensory conflict in VR?
- use steady velocity (for translations and rotations)
- no strafing (side-stepping movements, not natural)
- orient user, give landmarks (steady passive motion or fading in and out) (for teleportation)
- reduce view range (60 degree where information should be kept) (accommodation vs. convergence (visual strain))
definition of skeuomorphic
obsolete design elements which are retained for familiarity or out of tradition
definition of veridical
pertaining to an experience, perception or interpretation that accurately represents reality
What is world and head knowledge and which characteristics do each of these have?
world: includes affordance and signifiers, also mappings
head: logical, physical, semantic, cultural, … experience
Which 4 kinds of constraints can be defined?
- Physical
- Cultural
- Semantic
- Logical
-> reducing chance of making a mistake
Physical constraints (also, legacy problem and resistance)
- properties of physical world
- > limiting number of actions
+ legacy problem; many devices using same standard, changing to impactful
+ resistance; avoiding expense of massive change
cultural constraints
- rules set by society (ex. about behavior)
- are likely to change over time
- differ over cultures
- > making universal convention hard
logical constraints
- relationships between spatial or functional layout of components
- > natural mappings
semantic constraints
- about meanings; rely on knowledge of situation and the world
- can change over time
Why is it important to consider cultural norms?
knowing and implementing them creates comfort and harmony
not knowing them leads to discomfort and confusion -> people won’t use application
Task analysis (human-centered design)
a usable design starts with careful observations of how tasks being supported are actually performed, followed by a design process that results in a good fit to the actual ways the task gets performed
activity-centered controls
-> everything needed at the same time is at one place
- Forcing functions, 2. interlocks, 3. lock-ins and 4. lockouts
- failure at one stage preventing next step from happening (safety engineering)
- guiding in a proper sequence
- perfect compatible actions
- preventing the entering of a “dangerous” place (need to be kept passive and should i annoy user)
Advantage and Disadvantage of conventions
+ provide valuable guidance for novel situations
- difficult to enact change
Is a change of a device (it’s usage) always good (obviously just if it’s beneficial)?
no, because people dislike adjusting to new things
-> merits of change need to strictly overweight difficulties if change
- > psychological perceptions (conceptual models) matter not the physical consistency
- > therefore, adjustments take time and need to be worth it
sound as signifiers (basic ideas?)
- can provide information on another way as over display
- trade off between its aid and its annoyance
- > if sound feedback is expected, silence is a problem
What helps to bring new things in the market? (comfort users)
- letting it look like the old
- > rather standardize than individualize to enable rising intuitiveness
