Data Visualization Best Practices
Create clear, effective figures and tables for your dissertation. From chart selection to APA formatting—elevate your visual communication.
"Above all else, show the data."
The goal of every visualization is to communicate data clearly. Everything else—colors, gridlines, decorations—should serve that purpose or be removed.
Choosing the Right Chart Type
Match your visualization to your data and message. Each chart type serves specific purposes.
Bar Charts
Comparing categories or groups
Examples:
- • Comparing means across groups
- • Frequency distributions
- • Likert scale responses
Tips:
- Start y-axis at zero for honest comparison
- Order bars logically (alphabetically, by value, or conceptually)
- Use horizontal bars for many categories or long labels
Line Charts
Showing trends over time or continuous relationships
Examples:
- • Longitudinal data
- • Time series
- • Pre-post comparisons across groups
Tips:
- Reserve for truly continuous data or time series
- Don't connect categorical data with lines
- Limit to 4-5 lines for readability
Scatter Plots
Showing relationships between two continuous variables
Examples:
- • Correlation visualization
- • Regression analysis
- • Identifying outliers
Tips:
- Add regression line when appropriate
- Use jittering for overlapping points
- Consider adding marginal distributions
Pie/Donut Charts
Showing parts of a whole (use sparingly)
Examples:
- • Demographic composition
- • Budget allocation
- • Simple proportions
Tips:
- Use only for parts that sum to 100%
- Limit to 5-6 slices maximum
- Consider bar charts for easier comparison
- Start at 12 o'clock, largest slice first
Box Plots
Showing distribution characteristics and outliers
Examples:
- • Comparing distributions across groups
- • Identifying outliers
- • Showing variance
Tips:
- Excellent for comparing multiple groups
- Always explain components to readers
- Consider violin plots for more detail
Heat Maps
Showing patterns in matrices or correlations
Examples:
- • Correlation matrices
- • Cross-tabulations
- • Temporal patterns
Tips:
- Use colorblind-friendly palettes
- Order rows/columns meaningfully
- Include numerical values for precision
Table Design Principles
Well-designed tables communicate complex information efficiently. Follow these principles for clarity.
Purpose
Every table should answer a specific question. If it doesn't serve analysis or support arguments, remove it.
Example: Bad: 'Table of all variables'. Good: 'Descriptive Statistics for Key Study Variables'
Structure
Organize rows and columns logically. Group related items. Use clear, concise headers.
Example: Put dependent variables in rows, groups/conditions in columns for easy comparison
Formatting
Align numbers by decimal point. Use consistent decimal places. Minimize lines and borders.
Example: Use 2 decimal places for most statistics, 3 for p-values. No vertical lines.
Labeling
Include descriptive title, clear labels, units of measurement, and notes for abbreviations.
Example: Note: *p < .05, **p < .01, ***p < .001. CI = confidence interval.
APA 7th Edition Formatting
Follow these formatting requirements for figures and tables in APA style.
Common Visualization Mistakes
Using 3D effects on charts
3D distorts proportions and hinders comparison. Always use 2D charts—they're clearer and more honest.
Truncating the y-axis to exaggerate differences
Start bar charts at zero. If differences are meaningful, they'll be visible. Break the axis only when necessary and clearly indicate the break.
Using pie charts for comparisons
Humans compare lengths better than angles. Use bar charts when comparing categories. Reserve pie charts for simple part-to-whole relationships.
Too much color without purpose
Use color to encode meaning (groups, values, emphasis). Gray is your friend—use it for secondary information.
Insufficient labels or missing legends
Every axis needs a label with units. Every figure needs a clear legend. Readers shouldn't have to guess what symbols mean.
Low-resolution or pixelated images
Export at 300+ DPI for print. Use vector formats (PDF, SVG) when possible. Test figures at actual print size.
Frequently Asked Questions
Need Help with Your Data Presentation?
Our data analysis experts can help you create publication-quality figures and tables that clearly communicate your research findings.
Related Resources
Explore more resources in our Data Analysis collection