📖 Note 1: Final Exam Summary (Introduction)
Most Important Points
- Definition: Visual representation of data.
- Goals:
Show data, compare values, avoid distortion.
- Three Purposes:
Record, Reason, Communicate.
- Anscombe:
Same statistics ≠ Same data.
- Challenger:
Poor visualization can cause poor decisions.
- John Snow:
Visualization reveals patterns.
- Nightingale:
Visualization persuades people.
📖 Note 2: Final Exam Summary (Data & Image Models)
Most Important Points
- N: Nominal = Categories
- O: Ordinal = Ordered Categories
- Q: Quantitative = Numeric Values
- Dimensions: Describe data
- Measures: Analyze data
- Tidy Data:
Variable = Column, Observation = Row
- Bertin:
Position is the strongest visual channel.
- Expressiveness:
Show correct facts.
- Effectiveness:
Easy to perceive.
📖 Note 3: Final Exam Summary (Visual Perception)
Most Important Points
- Gestalt Laws: Explain how humans naturally organize visual information.
- Similarity: Similar objects form groups.
- Proximity: Close objects form groups.
- Closure: The brain fills missing gaps.
- Continuity: The eye follows smooth paths.
- Connectedness: Connected objects belong together.
- Figure-Ground: Separate important information from background.
- Prägnanz: Simpler visualizations are easier to understand.
- Preattentive Processing: Visual features are detected instantly.
- Preattentive Attributes: Color, Size, Shape, Orientation, Position.
📖 Note 4: Final Exam Summary (Graphical Integrity & Design)
Most Important Points
- Graphical Excellence: Give greatest ideas in shortest time with least ink.
- Graphical Integrity: Tell the truth about data (not lying with statistics).
- Lie Factor: Should equal 1.0 (accurate representation).
- Avoid Distortion: Visual representation proportional to numerical values.
- Avoid Ambiguity: Use clear labels and detailed annotations.
- Show Data Variation: Not design variation (avoid decoration).
- Account for Inflation: Adjust monetary values when comparing across time.
- Match Dimensions: Don't exceed data dimensions with visual dimensions.
- Show Context: Graphics should not quote data out of context.
- Data-Ink Ratio: Maximize data ink, minimize non-data ink (chartjunk).
- Small Multiples: Multiple similar charts placed together for comparison.
- KISS Principle: Keep visualizations simple.
📖 Note 5: Final Exam Summary (Visualization Design & Chart Types)
Most Important Points
- Form Follows Function: Choose design based on purpose.
- Affordances: Guide audience attention.
- Accessibility: Keep visualizations readable and simple.
- Shaffer's 4 C's: Clear, Clean, Concise, Captivating.
- Bar Charts: Start at zero.
- Pie Charts: Only for part-to-whole relationships.
- Scatterplots: Show relationships.
- Line Charts: Show trends over time.
- Boxplots: Compare distributions.
- Maps & Choropleths: Visualize geographic data.
📖 Note 6: Final Exam Summary (Dashboard & Storytelling)
Most Important Points
- Dashboard: Visual display of important information on a single screen.
- Dashboard Types: Strategic (goals), Tactical (projects), Operational (real-time), Analytical (exploration).
- Strategic Dashboards: Big-picture organizational goals, updated daily/weekly/monthly.
- Operational Dashboards: Real-time monitoring, immediate action support.
- Analytical Dashboards: Deep data exploration, discover insights and root causes.
- Data Storytelling: Data + Narrative + Visuals = complete story.
- Story Structure: Attract → Engage → Punchline (A-E-P).
- Laws of Attraction: Theme, audience, purpose, tone, simplicity.
- Rules of Engagement: Use visuals, color, layout and guidance effectively.
- Punchline: Deliver a clear and memorable conclusion (never leave guessing).
📖 Note 7: Final Exam Summary (Exploratory Data Analysis)
Most Important Points
- EDA: Explore data before visualization.
- Tukey (1977): Introduced EDA.
- EDA Goals: Find distributions, outliers, correlations and quality issues.
- Taxonomy of Interaction: Data & View Specification, View Manipulation, Process & Provenance.
- Selection: Point selection and region selection.
- Brushing & Linking: Highlight related data across views.
- Dynamic Queries: Interactive filtering with immediate feedback.
- Direct Manipulation: Pointing instead of typing (easier than text queries).
- Trellis Displays: Compare categories using multiple panels.
- Big Data: Use sampling, binning, modeling and aggregation techniques.
📖 Note 8: Final Exam Summary (Motion & Animation)
Most Important Points
- Motion: Directs attention and explains change.
- Common Fate: Objects moving together are grouped together.
- Animation Principles: Congruence and Apprehension.
- Three Uses: Narrative, Encoding, Transition.
- Transition Types: View, Representation, Surface, Time, Structure, Reordering, Filtering.
- 10 Principles: Ensure animations remain understandable (Congruence and Apprehension rules).
- Animation Tradeoff: Can improve understanding or become chart junk. Use only when beneficial.
- IBM Design Guideline: Animations should be UEICC - Understandable, Essential, Impactful, Consistent, Contextual.
📖 Bonus: Key Visual Encoding Principles (from Notes 2)
Tufte's Six Principles of Graphical Integrity
- Avoid distortion (Lie Factor = 1)
- Use clear labels
- Show data variation, not design variation
- Adjust monetary values for inflation
- Match graphic dimensions to data dimensions
- Provide context
Tufte's Five Laws of Data-Ink
- Above all else, show the data
- Maximize the data-ink ratio
- Erase non-data ink
- Erase redundant data ink
- Revise and edit
Bertin's Visual Variables
- Position: Quantitative (strongest encoding)
- Size: Quantitative
- Hue: Nominal
- Value: Ordinal
- Shape: Nominal