Unit 3: Designing effective visualizations
Data Visualization notes · PTU syllabus (UGDSE103)
On this page
Unit summary
Effective visualisations follow design principles: they are accurate, clear and focused. This unit covers principles of good design, effective use of colour, and why data modelling matters for building reliable visuals.
After this unit you can
- Apply the principles of good visualisation design
- Choose colour palettes effectively and accessibly
- Explain the role of data modelling in visualisation
- Critique and improve a poor chart
PTU syllabus topics
- Principles of good visualization design
- using color effectively
- the importance of data modeling in visualization
- Clarity
- One message per chart
- Colour
- Use colour to highlight, not decorate
- Honest axes
- Bars start at zero
- Labels
- Title, axis names and units
- Less clutter
- Remove gridlines and 3D effects
Topic 1
Principles of good visualisation design
Accuracy
Honest scales, axes from zero for bars
Clarity
One message per chart, clear titles and labels
Simplicity
Remove chart junk: heavy gridlines, 3-D, shadows
Data-ink ratio
Most ink should show data (Edward Tufte)
Right chart
Match the chart to the question
Consistency
Same colours and scales across charts
Topic 2
Using colour effectively
| Palette type | Use | Example |
|---|---|---|
| Sequential | Ordered values from low to high | Light to dark green for sales |
| Diverging | Values above and below a midpoint | Red–white–blue for profit and loss |
| Categorical (qualitative) | Distinct groups | Different colours for each course |
- Use colour to highlight what matters and grey for the rest.
- Limit to about 5–7 distinct colours.
- Ensure contrast and use colour-blind-safe palettes; don't rely on red versus green alone.
Topic 3
The importance of data modelling
Data modelling organises data into tables and relationships before visualising — for example a star schema with a fact table (sales) linked to dimension tables (date, product, region).
- Correct relationships ensure totals and filters work correctly.
- Calculated measures (for example profit margin) are defined once and reused.
- A clean model makes dashboards faster and easier to maintain.
Example
Without a relationship between Sales and Product tables, a "sales by category" chart may show the same total for every category.
Topic 4
Improving a poor chart
Type
3-D pie with 12 slices
Sorted horizontal bar chart
Labels
Legend far from data
Direct labels on bars
Colour
Rainbow colours
One colour, highlight the key bar
Axis
Truncated at 80
Starts at 0
Key terms
- Chart junk
- Decoration that adds no information
- Data-ink ratio
- Share of a chart's ink used to show data
- Sequential palette
- Colours showing ordered values
- Diverging palette
- Colours showing values on both sides of a midpoint
- Star schema
- A fact table connected to dimension tables
Quick revision
- Accurate, clear, simple, right chart, consistent.
- Sequential, diverging and categorical palettes.
- Highlight with colour, mute the rest; be colour-blind safe.
- Model data (star schema) before building dashboards.
Important exam questions
Practice questions written to the PTU exam pattern for this unit's syllabus: short answers (Section A style) and long answers (Sections B and C style).
Short-answer questions
- Q1.What is chart junk?
- Q2.Define the data-ink ratio.
- Q3.When should a diverging colour palette be used?
- Q4.Why is data modelling important for visualisation?
- Q5.Give two tips for colour-blind-friendly charts.
Long-answer questions
- Q1.Explain the principles of good visualisation design with examples.
- Q2.Explain how colour should be used effectively in visualisation.
- Q3.Discuss the role of data modelling in building reliable dashboards.
Stuck on this unit?
Message SBS on WhatsApp for help with Data Visualization, or to ask about studying BCA at Synetic.