Principles of Visual Design

Effective data visualisation is not just about choosing the right chart — it is about applying design principles that guide the viewer's eye, reduce cognitive effort, and ensure the message is understood quickly and accurately. This lesson covers Gestalt principles, colour theory, typography, accessibility, and the concept of data-ink ratio.

---

Gestalt Principles of Perception

The Gestalt principles describe how humans naturally group and organise visual elements. Understanding these principles helps you design charts that are intuitive to read.

Principle	Description	Application in Charts
Proximity	Elements close together are perceived as a group	Group related labels near their data points
Similarity	Elements that look alike are perceived as related	Use consistent colours for the same category
Enclosure	Elements enclosed in a boundary are perceived as a group	Use background shading to group chart regions
Continuity	The eye follows smooth paths	Align chart elements along axes and gridlines
Closure	The mind completes incomplete shapes	Minimalist chart borders still look complete
Connection	Connected elements are perceived as related	Lines connecting data points imply sequence

---

Colour Theory for Data Visualisation

Colour Models

Model	Description	Use
RGB	Red, Green, Blue — additive model for screens	Digital displays
CMYK	Cyan, Magenta, Yellow, Key (Black) — subtractive model	Print
HSL	Hue, Saturation, Lightness	Intuitive for choosing colours

Types of Colour Palettes

Palette Type	Description	When to Use
Sequential	Light to dark shades of one hue	Ordered data (e.g., temperature, revenue)
Diverging	Two hues meeting at a neutral midpoint	Data with a meaningful centre (e.g., profit/loss)
Qualitative	Distinct hues with similar lightness	Categorical data (e.g., countries, product types)

Common Colour Palettes in Python

import seaborn as sns
import matplotlib.pyplot as plt

# Sequential palette
sns.palplot(sns.color_palette("Blues", 8))
plt.title("Sequential: Blues")
plt.show()

# Diverging palette
sns.palplot(sns.color_palette("RdBu", 8))
plt.title("Diverging: Red-Blue")
plt.show()

# Qualitative palette
sns.palplot(sns.color_palette("Set2", 8))
plt.title("Qualitative: Set2")
plt.show()

Colour Dos and Don'ts

Do	Don't
Use colour purposefully to encode data	Use colour decoratively without meaning
Limit to 5-7 distinct colours per chart	Use a rainbow palette for sequential data
Test in greyscale to check readability	Rely solely on colour to convey information
Use a consistent palette across related charts	Change colour meanings between charts

---

Accessibility in Data Visualisation

Approximately 8% of men and 0.5% of women have some form of colour vision deficiency. Designing for accessibility is not optional — it is essential.

Guidelines for Accessible Charts