CSE 578 Exam #2 with complete updated
questions & answers
(1) Strive for Consistency
- Use similar action sequences and terminology (e.g. all car AC controls)
- Use consistent color, layout, and fonts (families, sizing, etc.)
(2) Enable frequent and easy use of shortcuts
- Be sure to consider different classes of users (e.g. novice users, normal
users, power users)
- As usage increases, so does desire to streamline common tasks.
3D Techniques
Space-time cube
- Spatial position is x-y, time is z
3D charts
- Deck.gl is an open source vis library developed at Uber
Key question: You are using low-accuracy channels, so do you really
need to use 3D?
,Low-accuracy channels: This suggests that the 3D channels being used,
such as depth perception or perspective, can be less accurate than 2D
channels due to issues like occlusion (where closer objects block the
view of objects behind them) or perspective distortion.
Might need 3D for time as a valuable data point
(3) Offer informative feedback
- For every user action, there should be an appropriate system feedback
- See also: responsiveness is required
(4) Design dialogues to yield closure
Sequences of actions should have a beginning, middle, and end. (e.g.
amazon buying sequence)
(5) Offer simple error handling
,- Try not to let the user make an error
- If they make an error, catch and handle it
- Error detection, explanation, and recovery
(6) Permit easy reversal of actions
- Undo, go back, revert, etc.
- to relieve anxiety and encourage exploration, because the user knows
he/she can always go back to previous states.
- Semantics: single-step undo? a data entry? a group of actions?
(7) Support internal locus of control
- If a person has an internal locus of control, that person attributes
success to their own efforts/abilities
- Make user action initiator, not respondent
- Avoid surprising system actions and difficulty with getting information
(8) Reduce short-term memory load
, - Human information processing is limited (e.g. working memory holds
7+-2 items, change blindness, external representation)
- To do so:
(1) keep interactions simple,
(2) have a clear structure for your app,
(3) reduce gulfs of execution and evaluation,
(4) leverage affordances and constraints OR provide instructions
Adjacency Matrix
From the network data (nodes + links), derive an adjacency matrix
- Tabular dataset with two key attributes (two lists of nodes)
- One value attribute for each ell to show a link between nodes
Adjacency Matrix - Drawbacks
Bad for path-based topology tasks (It's difficult to trace paths through
the graph using an adjacency matrix, as this requires following a
sequence of cells across rows and columns. This is more intuitive in
node-link diagrams)
Adjacency Matrix - Ordering
questions & answers
(1) Strive for Consistency
- Use similar action sequences and terminology (e.g. all car AC controls)
- Use consistent color, layout, and fonts (families, sizing, etc.)
(2) Enable frequent and easy use of shortcuts
- Be sure to consider different classes of users (e.g. novice users, normal
users, power users)
- As usage increases, so does desire to streamline common tasks.
3D Techniques
Space-time cube
- Spatial position is x-y, time is z
3D charts
- Deck.gl is an open source vis library developed at Uber
Key question: You are using low-accuracy channels, so do you really
need to use 3D?
,Low-accuracy channels: This suggests that the 3D channels being used,
such as depth perception or perspective, can be less accurate than 2D
channels due to issues like occlusion (where closer objects block the
view of objects behind them) or perspective distortion.
Might need 3D for time as a valuable data point
(3) Offer informative feedback
- For every user action, there should be an appropriate system feedback
- See also: responsiveness is required
(4) Design dialogues to yield closure
Sequences of actions should have a beginning, middle, and end. (e.g.
amazon buying sequence)
(5) Offer simple error handling
,- Try not to let the user make an error
- If they make an error, catch and handle it
- Error detection, explanation, and recovery
(6) Permit easy reversal of actions
- Undo, go back, revert, etc.
- to relieve anxiety and encourage exploration, because the user knows
he/she can always go back to previous states.
- Semantics: single-step undo? a data entry? a group of actions?
(7) Support internal locus of control
- If a person has an internal locus of control, that person attributes
success to their own efforts/abilities
- Make user action initiator, not respondent
- Avoid surprising system actions and difficulty with getting information
(8) Reduce short-term memory load
, - Human information processing is limited (e.g. working memory holds
7+-2 items, change blindness, external representation)
- To do so:
(1) keep interactions simple,
(2) have a clear structure for your app,
(3) reduce gulfs of execution and evaluation,
(4) leverage affordances and constraints OR provide instructions
Adjacency Matrix
From the network data (nodes + links), derive an adjacency matrix
- Tabular dataset with two key attributes (two lists of nodes)
- One value attribute for each ell to show a link between nodes
Adjacency Matrix - Drawbacks
Bad for path-based topology tasks (It's difficult to trace paths through
the graph using an adjacency matrix, as this requires following a
sequence of cells across rows and columns. This is more intuitive in
node-link diagrams)
Adjacency Matrix - Ordering
