How to Describe Data and Trends in French
Describe rising, falling and stable figures in French using a fictional dataset, accurate percentage and percentage-point calculations, and explained answers.
To describe a trend in French, first name the measure, period and unit. Then say what changed, by how much and what the figures cannot explain. This lesson uses a completely fictional table so you can practise accurate language without mistaking an invented number for real-world evidence.
A small fictional dataset
Imagine a made-up neighbourhood library counting visits and a survey of its visitors. The years, totals and percentages below are invented solely for this lesson. Survey percentages describe the share of surveyed visitors who chose a fictional evening-opening option; the table gives no survey size and no causal explanation.
| Measure | 2023 | 2024 | 2025 |
|---|---|---|---|
| Recorded visits | 1,000 | 1,200 | 1,200 |
| Share choosing evening opening in survey | 30% | 35% | 40% |
| Workshops offered | 20 | 18 | 15 |
Describe direction before explaining it
- Rise: Le nombre de visites est passé de 1 000 en 2023 à 1 200 en 2024. Est passé de... à... names both endpoints.
- Stable: Le nombre de visites est resté stable entre 2024 et 2025, à 1 200. It is not correct to call the whole 2023-2025 period stable: it rose first.
- Fall: Le nombre d'ateliers proposés a diminué de 20 à 15 entre 2023 et 2025. This is five fewer workshops.
- Comparison: En 2025, la bibliothèque a enregistré autant de visites qu'en 2024, mais elle a proposé trois ateliers de moins. The two trends differ.
Three kinds of change that must not be mixed
Absolute change: Visits rose from 1,000 to 1,200 between 2023 and 2024. The difference is 200 visits: Le nombre de visites a augmenté de 200. Workshops fell from 20 to 15 over 2023-2025, a decrease of five workshops.
Relative change: Divide the change by the starting value. For visits, 200 / 1,000 = 0.20, so Le nombre de visites a augmenté de 20 % entre 2023 et 2024. For workshops, 5 / 20 = 0.25, so Le nombre d'ateliers a diminué de 25 % entre 2023 et 2025. A fall of five is not a fall of five percent. The OQLF distinguishes absolute and relative variation.
Percentage-point change: The survey share rose from 30% to 40% across 2023-2025. That is une hausse de 10 points de pourcentage. Its relative increase is 10 / 30 = one third, or approximately 33.3%. Saying simply une hausse de 10 % would be ambiguous or wrong for this comparison. Use points de pourcentage when subtracting two percentages directly, as in the OQLF terminology entry.
Write a cautious description, not an invented cause
Dans ce jeu de données fictif, les visites de la bibliothèque augmentent entre 2023 et 2024, puis restent au même niveau en 2025. La part des personnes interrogées qui choisissent l'ouverture en soirée progresse chaque année, de 30 % à 40 % au total. En revanche, le nombre d'ateliers proposés baisse de 20 à 15. Ces chiffres décrivent des évolutions différentes; ils ne permettent pas de savoir pourquoi les visites ont augmenté ni si les mêmes personnes ont répondu à l'enquête chaque année.
The final sentence is important. With no sample sizes, methods or context, you cannot infer that demand for evening hours caused more visits. The invented survey share also cannot tell us how many people chose that option.
Try four checks, then compare the answers
- 2024-2025 visits: Describe the change in French. Answer: Le nombre de visites est resté stable à 1 200. Absolute and relative changes are both zero.
- 2024-2025 workshops: Give absolute and relative changes. Answer: Le nombre d'ateliers a baissé de trois, soit environ 16,7 % par rapport aux 18 ateliers de 2024. Calculation: 3 / 18 = 0.1666...
- 2024-2025 survey share: State the change in points. Answer: La part est passée de 35 % à 40 %, une hausse de cinq points de pourcentage. If a relative comparison is required, 5 / 35 is about 14.3%, which is a different measure.
- Interpretation: Can the table prove that fewer workshops caused higher attendance? Answer: No. It shows simultaneous changes, not the reason for either. We would need information about visitors, programming and measurement before making a causal claim.
When writing about real data, name the original source, date, sample and unit. In French, la part describes a share, le nombre a count, and environ marks a rounded figure. Those distinctions make a short explanation more trustworthy than decorative vocabulary.