Saying what you found, and what you did not

By the end of this lesson you can write a finding that is clear about its limits.

Why it matters

Readers remember your words, not your spreadsheet. If the words claim more than the data shows, people make decisions on things nobody measured. Being clear about limits is what makes people trust you next time.

The idea in plain words

Imagine you taste one spoonful of soup and say, "This soup is salty." Fair. Now imagine saying, "All soup in this city is salty." You only tasted one spoonful from one pot.

A finding should be the same size as the evidence. Three habits keep it honest:

1. Use the exact words for what was measured. "28 of 60 sales were cash", not "customers prefer cash" (nobody asked them what they prefer).
2. Name the scope: who, where and when the data covers (lesson 3.6).
3. Say what the data cannot tell you, such as why, or what happens elsewhere.

Overclaim: a statement that goes further than the evidence, for example to other places, other times, or to reasons that were never measured.

Limitation: something the data cannot show, which the reader needs to know.

Each honest version says only what was counted, where and when.

See it done

Draft finding: "Lagos customers love cash and will keep paying that way."

Data: market-stall-week.csv (synthetic).

1. Underline each claim. "Lagos customers", "love cash", "will keep paying that way".
2. Check each against the data. - "Lagos customers": the data covers one stall, not all of Lagos. Overclaim. - "love cash": the data shows how they paid, not what they love. Overclaim (a "why" question, lesson 3.6). - "will keep paying that way": the data covers one week in the past. It says nothing about the future. Overclaim.
3. Find what the data does show. =COUNTIF(H:H,"Cash") gives 28 of 60 sales, more than transfer (20) or card (12).
4. Rewrite at the right size. "At Amara's stall, in the week of 2 to 7 March 2026, 28 of 60 sales were paid in cash, more than any other way."
5. Add the limits. "One week at one stall cannot tell us about other stalls, other weeks, or why customers chose cash."

The honest version is longer, but every word can be checked.

Useful honest words: "in this data", "in this week", "at this stall", "about", "most of these", "we cannot tell from this data whether ...".

Your turn, with help

Rewrite: "Pupils in Bogotá hate walking to school."

Hint: what did Lucía's survey actually count, about whom, and when? =COUNTIF(F:F,"Walk") tells you how many walk.

Show a model answer

"In Lucía's survey of 30 pupils in classes 5A and 5B (March 2026), 4 walked to school, fewer than came by bus (14) or car (8). The survey did not ask whether pupils like walking, and it covers only these two classes."

Your turn, alone

Rewrite each overclaim honestly. Add one limitation.

1. "Friday is always Amara's best day." (Friday had 12 sales; the data is one week.)
2. "Qatar's people are the worst polluters in the world." (Use lesson 5.2: Qatar had the highest CO₂ per person of our 17 countries in 2024.)
3. "Hot weather makes people buy coconuts." (Lesson 5.5: stall-days data.)

Check your answers

Answers will vary. Good examples:

1. "In the week of 2 to 7 March 2026, Friday was Amara's busiest day, with 12 sales. We have only one week, so we cannot say whether Fridays are always busiest."
2. "Of the 17 countries in our data, Qatar had the highest CO₂ per person in 2024: 41.3 tonnes. Our data does not include every country, and CO₂ per person describes a country's total shared out, not the actions of any one person."
3. "On Amara's 28 days, hotter days had more coconut sales. This is made-up data for teaching, built so that heat drives sales; real data would need more evidence before saying heat causes it."

Watch out

• Feelings words. "Love", "hate", "prefer" claim to know minds. Data about actions shows actions.
• "Always" and "never". A week, or a year, cannot prove "always".
• Hiding the limits at the end in small print. Put the main limitation right next to the finding.

Say it back

1. What is an overclaim?
2. Name the three habits of an honest finding.
3. Why is "customers prefer cash" an overclaim for Amara's data?

Show the answers

1. A statement that goes further than the evidence.
2. Use the exact words for what was measured; name the scope; say what the data cannot tell you.
3. The data shows how customers paid, not what they prefer.

What is next

You can now write a data story that is clear, checked and honest. One lesson left: what you can do now, and where you can go from here.

---

Data in this lesson: Global Carbon Budget (2025); UN World Population Prospects (2024), processed by Our World in Data. Licensed under CC BY 4.0 (UN data: CC BY 3.0 IGO). Extract and continent labels by DataMedics.