Module 5 Lesson: Reading Charts and Distributions
Classroom Explanation
A chart encodes numbers as lengths, positions or areas, and readers decode them automatically. That decoding assumes things the chart does not always say — most often that bar length is measured from zero.
In this module, the goal is to read charts by separating what is shown from what is not proved. The important part is not only the calculation. The important part is the thinking before and after the calculation.
Why This Matters
The common mistake is copying the visible message of the chart without checking axis, denominator, sample size, and missing context. This mistake can make a report sound more confident than the evidence deserves.
Key Ideas
1. A chart should match the question and data type.
A chart is useful only when it fits the question. Counts, distributions, relationships, and time patterns usually need different visual forms.
2. Axes, scale, sample size, and grouping can change the message.
Small design choices can change a chart's message. Axis range, grouping, missing denominators, and sample size can make a weak pattern look strong.
3. A chart can show a pattern without proving the reason for that pattern.
A chart is an argument in visual form. Good readers check the scale, sample, and missing context before accepting the message.
Worked Example
A line chart may show signups rising after a campaign, but without a comparison group we cannot be sure the campaign caused the rise.
Here is the dataset used in this module.
| chart_id | chart_type | intended_message | sample_size | axis_risk | safe_observation |
|---|---|---|---|---|---|
| CH01 | bar chart | Course B has more completions than Course A | 80 | y-axis starts above zero | Course B appears higher but the axis may exaggerate the difference. |
| CH02 | line chart | Weekly signups increased after a campaign | 12 | no comparison group | Signups rose after the campaign but other causes are still possible. |
| CH03 | histogram | Most delivery times are under 45 minutes | 120 | outlier hidden by wide bins | Most times are below 45 minutes but bin width may hide long delays. |
| CH04 | scatter plot | Study hours and score move together | 60 | no confounder shown | Higher study hours are associated with higher scores in this sample. |
| CH05 | box plot | Group A has a lower median wait time | 45 | small sample per group | Group A's median appears lower but sample size should be checked. |
Check the axes before the shape. A truncated axis is not a lie about any individual value; it changes the ratio the reader perceives between them, which is why a six per cent difference can be drawn to look like a fourfold one with every number correct.
How To Think Through It
- Read both axes, including where they start.
- Compute the real difference as a proportion.
- Compare that with what the chart appears to show.
- Decide whether the axis choice is defensible.
- Write what the chart cannot show into the caption.
Common Mistake
The common mistake is copying the visible message of the chart without checking axis, denominator, sample size, and missing context.
How To Write The Result
The chart shows an increase after the campaign, but other causes are still possible because no comparison group is shown.
Practice Prompt
Pick three chart scenarios and write one observation plus one caution for each.
Takeaway
The axis is where charts mislead, and no individual number has to be wrong for it to happen.