Business Analytics 101: Data Visualization and Storytelling Principles of Visual Perception Preattentive Attributes Gestalt Principles

What This Is

Visual perception is the process by which our brains interpret visual information from the world around us. In business analytics, understanding visual perception is crucial for creating effective data visualizations that communicate insights and drive decision-making. For example, a retail company wants to identify patterns in customer purchasing behavior to inform marketing campaigns. By applying principles of visual perception, the company can create a dashboard that effectively communicates these insights to stakeholders, leading to better decision-making and improved business outcomes.

Key Formulas & Metrics

• Pre-attentive Attributes:
• Color Contrast Ratio = (L1 + 0.05) / (L2 + 0.05) – measures the difference in lightness between two colors.
• Color Hue = (H1 + H2) / 2 – average of two color hues.
• Size Ratio = (S1 + S2) / 2 – average of two object sizes.
• Position = (X1 + X2) / 2 – average of two object positions.

• Orientation = (θ1 + θ2) / 2 – average of two object orientations.

• Gestalt Principles:

• Proximity = (d1 + d2) / 2 – average distance between two objects.
• Similarity = (s1 + s2) / 2 – average similarity between two objects.
• Continuity = (c1 + c2) / 2 – average continuity between two objects.
• Closure = (c1 + c2) / 2 – average closure between two objects.
• Figure-Ground = (f1 + f2) / 2 – average figure-ground relationship between two objects.

Step-by-Step Procedure

1. Identify the goal: Determine the purpose of the data visualization (e.g., to communicate a trend, to highlight a pattern).
2. Choose the right chart: Select a chart type that effectively communicates the goal (e.g., line chart for trends, bar chart for comparisons).
3. Apply pre-attentive attributes: Use color, size, position, and orientation to draw attention to key insights.
4. Apply Gestalt principles: Use proximity, similarity, continuity, closure, and figure-ground to create a clear and cohesive visual narrative.
5. Test and refine: Test the visualization with stakeholders and refine it based on feedback.
6. Document and share: Document the visualization and share it with stakeholders to ensure effective communication.

Common Mistakes

• Mistake: Confusing correlation with causation.
• Correction: Establish a clear cause-and-effect relationship between variables before drawing conclusions.
• Mistake: Misinterpreting p-values.
• Correction: Understand the p-value as a measure of evidence against the null hypothesis, not as a probability of the null hypothesis being true.
• Mistake: Using the wrong error metric for a business problem.
• Correction: Choose an error metric that aligns with the business objective (e.g., mean absolute error for forecasting, precision for classification).

Software / Tool Tips

• Tableau: Use the "Color" pane to apply pre-attentive attributes and the "Layout" pane to apply Gestalt principles.
• Power BI: Use the "Conditional Formatting" feature to apply pre-attentive attributes and the "Layout" feature to apply Gestalt principles.
• Python with pandas and scikit-learn: Use the "matplotlib" library to create visualizations and apply pre-attentive attributes and Gestalt principles.

Quick Practice Problem

Scenario: A company wants to create a dashboard to communicate sales trends to stakeholders. The dashboard should highlight the top-performing product category.

Question: What does an R² of 0.85 mean?

Answer: An R² of 0.85 means that 85% of the variance in sales is explained by the product category.

Last-Minute Cram Sheet

• Pre-attentive attributes: Color, size, position, orientation, and movement.
• Gestalt principles: Proximity, similarity, continuity, closure, and figure-ground.
• R²: Measures the proportion of variance explained by a regression model.
• p-value: Measures the evidence against the null hypothesis.
• Mean Absolute Error (MAE): Measures the average absolute difference between forecast and actual values.
• ⚠️ p-value is NOT the probability that H₀ is true – it’s the probability of observing the data (or more extreme) if H₀ is true.
• ⚠️ Correlation does NOT imply causation.
• ⚠️ R² is NOT a measure of goodness of fit – it’s a measure of variance explained.