Chapter 5 · 7 levels
Choosing and reading charts
Which chart answers which question, how a cut-off axis exaggerates a gap, when small multiples beat one busy chart, smoothing noisy daily data, and reviewing a dashboard chart honestly.
StartA chart is an argument. It decides what the reader compares, and readers trust what they see more than any table under it. That is why choosing and reading charts is an analyst skill, not a design one: the wrong chart can make a true number say something false.
In this chapter every chart is drawn from Harbor & Pine’s real data. You start by matching questions to chart types: change over time, comparing groups, the shape of one column, and the relationship between two. Then you measure how much a cut-off axis exaggerates a gap, see why six lines on one chart hide what six small charts show, smooth noisy daily revenue with a 7-day rolling average in pandas, show parts of a whole without a pie that hides the answer, plan a dashboard around the decision it supports, and finish by reviewing a dashboard chart that is technically correct and still misleading.
The code levels compute the numbers behind a chart, because fixing a chart usually starts with fixing what it plots.
Levels
- Which chart answers the question?
from Thomas Whitaker, Head of Merchandising
chart choice, line chart, bar chart, histogram, scatter plot
- The bar that looks four times taller
from Daniel Foster, Mobile Product Manager
truncated axis, bar chart, conversion rate, misleading charts
- Six lines or six small charts
from Jessica Turner, Merchandising Manager
small multiples, line chart, seasonality, chart choice
- Smooth the daily revenue chart
from Laura Bennett, Head of Finance
rolling average, resample, time series, smoothing
- Parts of a whole, without the pie
from Thomas Whitaker, Head of Merchandising
pie chart, shares, stacked column, dual axis, small multiples
- Design the CEO dashboard
from Catherine Hale, Chief Executive Officer
dashboard design, kpis, data storytelling, audience
- Review the momentum chart
from Jordan Reyes, Head of Marketing
cumulative chart, year over year, misleading charts, chart review