Interpret results without hiding uncertainty · Interpréter les résultats sans dissimuler l'incertitude
| English | Français |
|---|---|
| analysis/əˈnæləsɪs/ | analyse |
| variable/ˈveərɪəbl/ | variable |
| limitation/ˌlɪmɪˈteɪʃn/ | limitation |
Two sites differ, but what does that mean?
- A student measures warmer ground in an open garden than under a tree. The result may fit a shade explanation, but the sites also have different surfaces. A number alone does not tell us which factor caused the difference.
- Analysis · Analyse 分析 explains patterns in evidence and tests what they mean for the research question. Record the conditions before deciding what the comparison shows.
Keep the record traceable
- Keep the original observations, units, dates, instrument details and site descriptions. Show how any summary was calculated so a reader can follow the steps.
- Do not remove a value simply because it disagrees with your expectation. Check whether a recorded error explains it, and document any exclusion rule and its effect.
What should you do with an unexpected observation? · Que devez-vous faire face à une observation inattendue ?
Unexpected data require investigation, not silent alteration. · Les données inattendues nécessitent une investigation, pas une modification silencieuse.
Compare like with like
- A variable 变量 is a factor that can change. If shade and surface both differ between sites, your comparison cannot separate their effects.
- Where feasible and approved, compare the same surface with different shade conditions and record repeated observations at consistent times. Better comparisons reduce competing explanations; they do not make all uncertainty disappear.
Why does the paved-versus-soil comparison leave a competing explanation? · Pourquoi la comparaison entre surface pavée et sol nu laisse-t-elle une explication concurrente ?
A surface difference could contribute to the temperature difference attributed to shade. · Une différence de surface pourrait contribuer à la différence de température attribuée à l'ombrage.
More readings at mismatched sites automatically separate the effects of shade and surface. · Des relevés supplémentaires effectués dans des sites incompatibles ne séparent pas automatiquement les effets de l'ombrage et de la surface.
Repeating the same comparison does not isolate factors that change together. · Répéter la même comparaison ne permet pas d'isoler les facteurs qui varient ensemble.
Explain a limitation’s effect
- A limitation 局限 is a reason the evidence supports less than you would like. Explain whether it affects measurement accuracy, comparison or the range of places your conclusion can cover.
- “The sample is small” is incomplete. State what range of times or sites it misses and how that prevents a wider claim. Suggest a realistic improvement that addresses the same issue.
Which limitation statement explains its effect? · Quelle déclaration de limitation explique son effet ?
The statement identifies a missing range and limits the conclusion accordingly. · L'énoncé identifie une plage manquante et limite la conclusion en conséquence.
Answer the question you investigated
- A defensible conclusion states the observed pattern, what explanation is supported and which competing explanation remains. Keep its scope tied to the actual method.
- If evidence is insufficient for the original question, explain the gap honestly and discuss the next step with your supervisor. Confirm the subject’s permitted method and the applicable assessment guide.
Match each record to its purpose. · Reliez chaque enregistrement à son objectif.
A reader needs both the results and the conditions that produced them. · Un lecteur a besoin à la fois des résultats et des conditions qui les ont produits.
Which conclusions fit the evidence described? · Quelles conclusions s'alignent avec les preuves décrites ?
The observed local pattern is supported, but the comparison does not isolate shade. · Le modèle local observé est étayé, mais la comparaison n'isole pas l'ombrage.
The open site is paved; the shaded site has soil. Three measurements at each site show a difference, but surface and shade changed together. The student can report that these two sites differed under the recorded conditions. To test a shade explanation more directly, an approved follow-up could compare the same surface in shaded and unshaded conditions, with consistent timing and instruments. More readings at the original mismatched sites would improve the local record without separating the two factors.
Do not turn a list of weaknesses into a conclusion. Explain which weakness changes the claim and how.
Keep a traceable record, check the comparison, explain material limits and answer only what your evidence supports.