Evidence review
Track claim band, source, limitation and action.

Match claim strength to evidence, limits and audience—then make a recommendation you can defend.
For learners: Align evidence quality, claim strength, scope and recommendation strength so decisions remain clear and proportionate.
For teachers: C1 progression from evidence–claim separation through 12 calibrated statements, precise scope and four repairs to an independent evidence-based recommendation, challenge and focused retry.
สำหรับผู้เรียน: ทำให้คุณภาพหลักฐาน น้ำหนักข้อความ ขอบเขต และความแรงของข้อเสนอแนะสัมพันธ์กัน เพื่อสื่อสารการตัดสินใจอย่างชัดเจนและพอเหมาะ
สำหรับครู: เส้นทาง C1 เริ่มจากแยกหลักฐานกับข้อกล่าวอ้าง ปรับน้ำหนัก 12 ข้อ กำหนดขอบเขต แก้ 4 ข้อ และจบด้วยข้อเสนอแนะจากหลักฐาน การตอบคำท้าทาย และการลองใหม่เฉพาะจุด
Separate evidence from claim strength, calibrate twelve statements, repair four vague or overstated positions, and present a balanced evidence-based recommendation.
แยกสิ่งที่หลักฐานพิสูจน์ได้ออกจากระดับความมั่นใจของประโยค ปรับน้ำหนักข้อความ 12 ข้อ แก้ข้อความที่คลุมเครือหรือแรงเกินหลักฐาน 4 ข้อ และนำเสนอข้อเสนอแนะที่สมดุลโดยอ้างอิงหลักฐานและข้อจำกัดอย่างชัดเจน
A hedge is not simply “weak language,” and a booster is not automatically confident or persuasive. At C1, the speaker must align certainty, scope and recommendation strength with the quality of evidence. Too little commitment can hide a useful conclusion; too much can turn a limited finding into a false fact.
ภาษาไทย: Hedge ไม่ได้แปลว่าพูดอ่อนเสมอไป และ booster ก็ไม่ได้แปลว่าพูดดีหรือมั่นใจเสมอ ผู้เรียนระดับ C1 ต้องทำให้น้ำหนักของประโยค ขอบเขต และข้อเสนอแนะสัมพันธ์กับคุณภาพหลักฐานจริง
Evidence before language. Establish what the source supports, for whom and under which conditions before selecting a hedge or booster. Do not teach caution as an automatic collection of maybe, possibly and seems.
| Stage | Teacher move | Evidence to collect |
|---|---|---|
| 1 · Separate | Map evidence, inference and unsupported leap. | 6 evidence maps |
| 2 · Scale | Calibrate 12 statements across four strength bands. | 12 calibrated claims |
| 3 · Scope | Add population, condition, time and source boundaries. | 8 scoped claims |
| 4 · Repair | Replace vague hedging and unsupported boosting. | 4 justified repairs |
| 5 · Transfer | Audit calibration across two accents. | 2 evidence-strength maps |
| 6 · Perform | Score MEAN/FORM/USE and target one retry. | Recommendation + challenge + retry |
Decide what the evidence supports before changing the wording.
| Evidence note | Defensible claim | Unsupported leap |
|---|---|---|
| 8 of 10 participants improved in one short trial | The approach appears promising in this small trial. | The approach clearly works for everyone. |
| two reports show the same direction; measures differ | The findings broadly suggest a positive trend. | The evidence conclusively proves the effect. |
| one expert predicts a risk | The change may create a risk that merits monitoring. | The change will cause serious harm. |
Ask “What exactly did the source observe?” and “What did the writer add?” Do not accept a cautious modal if the rest of the sentence still generalises beyond the sample or method.
Choose the narrowest strength band that communicates the evidence honestly.
| Strength band | Language resources | Evidence condition |
|---|---|---|
| 1 · possibility | may · might · could · it is possible that | plausible but limited or indirect |
| 2 · tendency | appears to · seems to · tends to · is likely to | pattern present but not conclusive |
| 3 · supported conclusion | the evidence suggests/indicates · is strongly associated with | consistent support with known limits |
| 4 · strong confirmation | demonstrates · establishes · clearly shows | direct, robust and independently confirmed |
Clearly cannot create clarity that the evidence lacks. May possibly seem to does not add honest caution; it creates vague stacking. One well-chosen device plus an explicit limit is usually stronger communication.
Require a reason for every Band 3 or 4 choice. If learners choose only low-strength language, ask which findings are sufficiently supported to state more directly.
Make claims precise instead of merely cautious.
| Precision job | Useful frame | Model |
|---|---|---|
| limit population | among / for / within this sample | Within this sample, the change was associated with faster completion. |
| limit condition | when / provided that / under these conditions | The approach is likely to help when support is available. |
| name source distance | according to / the review suggests / self-reported | According to self-reported data, satisfaction appears to have increased. |
| separate result from recommendation | supports testing / does not yet justify full adoption | The result supports a limited pilot but does not yet justify a full rollout. |
Reward boundaries before adverbs. “In this six-week pilot” communicates more than “possibly.” Check that recommendation language remains a decision based on evidence, not a direct restatement of the result.
Fix strength, scope and recommendation without hiding behind soft language.
| Problem position | Calibration problem | Repair target |
|---|---|---|
| This clearly proves that remote work improves performance. | association becomes universal proof | The study suggests an improvement for… |
| The result may possibly perhaps indicate a benefit. | stacked hedges obscure commitment | The result indicates a possible benefit in… |
| There is some evidence, so we should definitely implement it everywhere. | limited evidence drives absolute action | supports a monitored pilot, not full adoption |
| Experts believe the method works. | source and degree are vague | name the review, finding and limitation |
Do not accept “add more hedges” as the repair strategy. The learner must align the strength band and make the limit visible. Challenge absolute recommendations that are unsupported even when the preceding claim sounds cautious.
Recover evidence strength, scope and recommendation from two C1 briefings.
Track claim band, source, limitation and action.
Track where the speaker strengthens, limits or separates a recommendation.
Audio never starts automatically. The two speakers provide voice transfer even within one accent variety; the task is calibration, not accent imitation. Replay only the line where evidence, strength, scope or action is lost.
Present a proportionate action and improve one precise calibration weakness.
Result: 0–2 Rebuild · 3–4 Developing · 5–6 C1 mastery evidence.
Score the first independent attempt across MEAN, FORM and USE. Give one target only: evidence–claim match, scope, strength device, action proportion or challenge response. Keep the original score and record the retry separately.
You can now calibrate a claim and a recommendation to the evidence behind them. Use the Dynamic Learning Graph below to revisit stance attribution, compare B2 caution strategies or continue to fine modal stance at C2.
Return to Grammar LessonsReporting Verbs for Stance
เรียนการรักษาท่าทีของแหล่งข้อมูลก่อน
Probability, Caution, and Hedging
เปรียบเทียบการสื่อความเป็นไปได้และความระมัดระวังระดับ B2
News: Fact, Opinion, and a Clear Summary
นำไปใช้แยกข้อเท็จจริง ความเห็น และสรุปข่าว
Fine Modal Stance and Interpersonal Distance
พัฒนาการควบคุม modal stance และระยะห่างระดับ C2