Measurement & Experimental Science
Data Visualization & Honest Cultivation Comparisons
Charts should reveal the underlying variation, sample size, time, and treatment structure instead of using dramatic axes or averages alone to make small differences look decisive.
Key concepts
- Raw points or distributions can reveal variation hidden by means
- Axis choices affect visual interpretation
- Time-series data should preserve event timing
- Missing data and excluded observations should be disclosed
What to measure or observe
- Plot individual values when sample size allows
- Show units and sample counts
- Use comparable axes for comparable treatments
- Mark interventions on time-series plots
Common mistakes
- Starting an axis at a value that exaggerates a trivial difference
- Showing only the best plant photo
- Hiding failed replicates
- Using a bar chart with no indication of variation
Visuals this lesson still needs
- Misleading-versus-honest axis
- Mean-only versus raw-points chart
- Annotated time-series
- Missing-data disclosure example