Skip to content
FORGENO ADS GAMESUnlock all $8

Outbreak Lab · Unit 10 of 11

Unit 10: Reading outbreak charts

Log and linear scales, data that arrives late, and comparing places fairly.

Opens the first lesson of this unit in Outbreak Lab. Every lesson is open from the start.

Lesson 1: Linear and log scales

On a linear scale, equal steps up the axis add the same amount: 0, 1,000, 2,000, 3,000. On a logarithmic (log) scale, equal steps multiply by the same amount: 10, 100, 1,000, 10,000.

Exponential growth looks like a hockey stick on a linear chart: flat, then suddenly steep. On a log chart it is a straight line, and the steepness of the line shows the growth rate. When the line bends flatter, growth is slowing, even if the linear chart still shoots upward.

Log charts are good for comparing growth rates and spotting changes. Linear charts are better for seeing actual sizes, like how many hospital beds are needed. Always check the axis before reading the shape.

Key idea: On a log scale, a straight line means a steady growth rate. Check the axis first.

Log scale
An axis where equal steps multiply by the same amount, such as 10, 100, 1,000.
Epidemic curve
A chart of new cases over time.

Try it in the Chart reader simulation: Switch between linear and log, then find the bend.

Lesson 2: Data arrives late

Case counts show when infections were reported, not when they happened. Between infection and report come the incubation period, the time to get tested and the time for results to be recorded. Today's reports often reflect infections from a week or more ago. Hospital admissions and deaths lag even further.

So the effect of a change, like a new measure, shows up in the data only after a delay. Judging a measure after just a few days is reading the past.

Reports also have rhythms: fewer tests and reports on weekends and holidays make dips that are not real. A 7-day average smooths these out.

Key idea: Reports lag infections, and deaths lag further. Use 7-day averages and allow for the lag.

Reporting lag
The delay between an infection and the day it appears in the data.
7-day average
The average of the last seven days, which smooths out weekly reporting patterns.

Try it in the Chart reader simulation: After you check, reveal the hidden infections line to see the lag.

Lesson 3: Comparing places fairly

1,000 cases means something very different in a town of 10,000 than in a city of 10 million. Rates per 100,000 people make places comparable.

Counts also depend on testing. More testing finds more cases, including mild ones. The share of tests that come back positive (test positivity) helps show whether testing is keeping up: high positivity suggests many infections are being missed.

Before trusting a chart, ask: what is counted, per how many people, on what scale, starting when, and how complete are the most recent days?

Key idea: Use rates per 100,000, watch for changes in testing, and treat the latest days as incomplete.

Rate per 100,000
Cases divided by population, times 100,000, so places of different sizes can be compared.
Test positivity
The share of tests that come back positive.

Every lesson is free: no ads, no lives and no paywall.