ShelfData AnalystData Scientist
Data visualisation and storytelling · Beginner · 3 h
Choosing the right chart
Match the chart to the question, and avoid the chart types that mislead.
Why employers care: Every analyst output ends in a chart. A wrong chart choice is the fastest way to lose a stakeholder's trust.
In the roadmaps for: Data Analyst, Data Scientist
Lessons, in order
- How to choose the right chart for your dataPick the chart from the question you are answering, know which visual encodings people read accurately, and avoid the five mistakes that make a correct number look wrong.12 min
- Bar chart best practices: sorting, baselines, labelsBuild bar charts that rank categories at a glance, choose correctly between stacked, grouped and 100% stacked bars, and handle long tails and negative values.12 min
- Line chart best practices for time series dataDraw time series that tell the truth about trend, seasonality and events, with honest smoothing, partial periods, indexed comparisons and no spaghetti.12 min
- Histograms, box plots and violin plots explainedChoose between a histogram, box plot, violin plot and ECDF, read each one correctly, and know what each can hide, using 3 million real taxi fares.11 min
- Scatter plots: show relationships and correlationBuild scatter plots that show a relationship honestly, from 200 countries to 3 million taxi trips, with log axes, bubble sizes, highlights and trend lines that do not overclaim.12 min