Training Data Visualization Chart Types & When to Use Them
1 / 10

Chart Types & When to Use Them

23 min Data Visualization

Chart Types & When to Use Them

Core Chart Types
  • Bar chart — compare categorical quantities
  • Line chart — show trends over time
  • Scatter plot — reveal relationships between two quantitative variables
  • Pie chart — show parts of a whole (use sparingly, ≤ 5 slices)
  • Histogram — display the distribution of a single quantitative variable
  • Box plot — summarize spread, center, and outliers
Choosing a Chart
GoalChart
Compare categoriesBar (horizontal if many labels)
Show trend over timeLine
Show distributionHistogram, box plot
Show relationshipScatter plot
Part-to-wholeStacked bar, pie
Heat / densityHeat map, contour
Data-Ink Ratio (Tufte)

Maximize the share of ink used to display data. Remove chart junk — unnecessary gridlines, 3-D effects, decorative elements.

Example 1

You have monthly revenue for 12 months. Which chart?

Line chart — reveals the trend across time.

Example 2

You want to compare exam scores in five classes. Which chart?

Side-by-side box plots — show medians, spreads, and outliers.

Example 3

Why is a pie chart bad for 12 categories?

Humans are poor at judging small angle differences — a bar chart makes comparisons far easier.

Practice Problems

1. Best chart for showing temperature over a week?
2. Best chart for showing market share of 4 companies?
3. What is chart junk?
4. When is a horizontal bar chart preferred?
5. Histogram vs. bar chart: what's the difference?
6. Best chart for height vs. weight of 100 students?
7. When would you use a heat map?
8. What does a box plot whisker represent?
9. Stacked bar vs. grouped bar: which for comparing within groups?
10. What is the data-ink ratio?
11. Best chart for showing a budget breakdown (5 items)?
12. Why avoid 3-D bar charts?
Show Answer Key

1. Line chart

2. Pie chart (or bar chart)

3. Non-data visual elements that clutter the chart

4. When category labels are long

5. Histogram: continuous data in bins; bar chart: discrete categories with gaps

6. Scatter plot

7. When showing values across two categorical dimensions (e.g., correlation matrix)

8. Range of non-outlier data (typically 1.5 × IQR from quartiles)

9. Grouped bar (individual bars are easier to compare)

10. Proportion of ink that represents actual data vs. total ink

11. Pie chart or stacked bar

12. Perspective distortion makes bar heights harder to compare accurately

Histograms & Distributions