Visual Perception and Data Visualization Principles

Gestalt Psychology and Its Influence on UI Design

• Gestalt Psychology: An early 20th-century study focusing on the organizing principles of vision. Humans inherently seek patterns, a concept that significantly aids in User Interface (UI) design. For further reading, many visualization books cover this topic extensively.
• Gestalt Psychology: Understanding these innate patterns helps direct attention and organize information effectively. Utilize color and spacing strategically for impactful design.

Psychophysical Laws in Perception

• Weber's Law: States that the just-noticeable difference between two stimuli is proportional to their magnitude. This indicates that human perception operates based on percentage increases.
• Steven's Power Law: Describes the non-linear relationship between actual and perceived value in perception. It is expressed as S=IN, where each perceptual channel has a different power, N.

Fundamental Graph Types for Data Visualization

Line Graph, Scatter Plot, Column (Bar), Dot-Chart (Horizontal Bar), Stacked Bar, Area, Bubble, Contingency Plot, Small Multiples.

Advanced Visualization Techniques

• Contingency Tables: These tables visualize relationships between two categorical variables, with at least one numerical variable. Categorical variables are represented on the axes, and numerical variables are depicted through markings like color and size. They are often derived from aggregated functions such as counts.
• Small Multiples: A generalized form of a contingency table, featuring categorical variables and various graphics arranged in a grid. They use consistent scales for simultaneous comparison and are also known as lattice plots.
• Common Node-Link Diagrams: These often feature an Inside-Out Ordering, with the root node in the center and leaf nodes on the outside.
• Tree Map: A hierarchy visualization that uses a containment metaphor, drawn recursively. It is a core feature in Tableau and is available as the "treemap" package in R.

Essential Graph Terminology

Data Region, Tick Marks, Axis, Data Labels, Plotting Symbol, Key, Reference Line.

Best Practices for Bar Charts

Bar charts require a zero baseline to accurately compare lengths, invite relative comparisons, and prevent distortion of proportional differences.

General Design Principles for Effective Visualization

Prioritize the viewer's eyes over memory. Emphasize side-by-side comparisons. Address change blindness. Offer multiple levels of detail. Ensure clarity in black and white. Prioritize function over form. Iteratively improve designs that are initially ugly but functional. Be cautious about reversing these priorities.

Using Grid Lines Effectively

Grid lines should not compete with the data. Use lighter colors, especially for minor lines, to aid judgment without obscuring the data.

Example of Clutter and Simplification

Clutter Example - To improve, simplify the Y-axis, improve X-axis labeling, move the legend, strengthen symbols, and remove error bars. This provides more room for uniform annotations like Rings or Janus.

Understanding Data Ink and Non-Data Ink

• Data Ink: Refers to the ink or pixels that display data and change with the data.
• Non-Data Ink: This is clutter-causing ink that diverts attention and occupies space meant for data.