Match claim strength to evidence by limiting scope, probability and causal commitment, while retaining confident language when the evidence genuinely warrants it.
Today’s central idea: Hedging calibrates commitment; it should not hide weak thinking. Limit the exact dimension the evidence cannot support: probability, frequency, quantity, population, cause, time or source. Use boosters when evidence and purpose justify strength.
Retrieval bridge
Say each answer before revealing it. The goal is to reactivate language that today’s lesson will extend.
Connect to what you already know
Soften: The policy caused the improvement.
The policy may have contributed to the improvement.
Limit a generalization about users.
In this sample, many users reported… / Some participants appeared to…
Attribute rather than endorse: The system is unsafe.
Critics argue that the system is unsafe.
Notice before the rule
Four claims from one small study
A pilot involved forty volunteers for two weeks.
The app improves sleep.
The app may improve sleep.
The results suggest that the app may improve sleep for some users.
Within this small sample, reported sleep improved slightly during the trial.
Which sentence makes the strongest general causal claim?
Which version limits population, evidence type and degree?
Does cautious wording necessarily mean the result is unimportant?
Noticing check
Which claim best matches a small short self-report study?
Discover the system
Hedging calibrates commitment; it should not hide weak thinking. Limit the exact dimension the evidence cannot support: probability, frequency, quantity, population, cause, time or source. Use boosters when evidence and purpose justify strength.
Pattern
Form
Example
modal/adverb
may, might, could; probably, apparently
The change may reduce waiting times.
reporting frame
suggests/appears/seems that
The evidence suggests that access improved.
scope limit
in this sample; for some users; under these conditions
In this sample, most users improved.
causal limit
contribute to; be associated with; coincide with
Training was associated with higher retention.
Meaning and choice
Speaker’s choice
Use it when…
Example
probability
the event or explanation is possible rather than certain
The delay may reflect staff shortages.
scope
the claim applies to a limited group, time or condition
Among new users, satisfaction rose.
attribution
a view belongs to a source
The authors argue that the policy is effective.
booster
strong evidence warrants commitment
The records clearly show that payment was received.
Sound natural: In speech, place a hedge early enough for the listener to interpret the whole claim: It MAY be that… Use a brief pause around scope limits: In THIS sample | the effect was small.
High-value repairs
Avoid
Use
Why
Hedges make every claim polite and accurate.
A hedge must target the actual uncertainty or scope limit.
Vague caution can still mislead.
The study proves the treatment might help.
The study suggests that the treatment might help.
Proves conflicts with might and overstates the evidence.
It is definitely possible that costs may perhaps rise.
Costs may rise.
Stacked hedges make the commitment unclear without adding precision.
Claim-strength controls
Identify which dimension needs limitation.
Likelihood
may/mightprobablyappears tois likely to
Scope
some/manyin this sampleunder these conditionsto a limited extent
Evidence relation
suggestindicatebe associated withcontribute to
Strength
clearlystrongly supportsdemonstratesis established by
Read language doing real work
A responsible pilot summary
Track every limit placed on the claim.
In a six-week pilot involving eighty employees, participants reported slightly lower stress and fewer missed deadlines. The results suggest that flexible start times may have helped some teams.
The pilot did not use a control group, and workload also fell during the period. The schedule change therefore cannot be treated as the sole cause of the improvement.
However, attendance records clearly show a 12 percent reduction in late arrivals. A larger controlled study would be needed to determine whether the pattern continues across departments.
Build controlled accuracy
Reusable practice routes
Choose today’s practice set
Choose the set that fits today’s learner. Each set contains different examples.
Core: Secure the central form-and-meaning decisions before increasing the pressure.
A correlation does not establish cause. Which claim fits?
A result occurred only among new users. Which scope limit fits?
A source holds a disputed view. Which attribution is accurate?
Records conclusively show payment. Which booster is justified?
Contrast: Use the reading context to distinguish nearby meanings and prevent overgeneralization.
Which population is covered?
Why is may have helped appropriate?
Which finding supports stronger wording?
What remains unknown?
Challenge: Apply the same system to less predictable contexts without repeating the Core examples.
Which sentence limits degree?
Which claim fits an observed timing relationship only?
Which revision avoids useless hedge stacking?
Which sentence accurately names self-reported evidence?
Core set selected.
Build the calibrated causal claim
Limit cause and population.
Diagnose and repair
Find the precise problem. A useful correction includes the reason, not only the repaired sentence.
Limit causal certainty
The pilot proved that flexible hours caused the improvement.
Name the population
Employees prefer flexible hours, according to eighty volunteers.
Remove hedge stacking
It is definitely possible that costs might perhaps rise.
Match reporting verb and stance
The report proves that the change may help.
Fluency drill
Respond aloud first. Reveal the model only after you have committed to an answer; your wording can differ if the meaning and form are right.
Transform and expand
State a strong claim, then calibrate probability, scope and cause.
The policy improved performance becomes In this sample, the policy may have contributed to a modest improvement in reported performance.
Attribute a disputed claim.
Supporters argue that the measure improves access, while the current data remain inconclusive.
Use a justified booster from direct records.
The payment records clearly show that the full amount was received on Tuesday.
Self-correct a generalization.
Users prefer the new design. More accurately, most participants in this small trial preferred it.
Play and practice
Games and a short reading that recycle this lesson. Use them as a warm-up, a fast finisher, or homework. Each one checks itself.
Read
Text reminders at the library
Read once for meaning. Then read again and do the tasks below.
Pilot summary245 wordsAbout 2 min
For eight weeks this spring, the Northside branch library sent text reminders to borrowers two days before their books were due. Each message included the title of the book and a link for renewing it online. Around three hundred people took part, all of whom had volunteered to join the trial. During the pilot, overdue returns fell from 14 percent to 9 percent of loans. The results suggest that reminders may reduce late returns, at least among borrowers who agree to receive them.
Two features of the trial limit what we can say. First, volunteers might be more organized than the average borrower, so the effect could be smaller across the whole library. Second, the branch extended its opening hours in April. Part of the improvement may therefore reflect easier access rather than the messages themselves.
Not every finding needs this much caution. The library’s own records clearly show that the text system cost less than $200 to run, and staff spent far less time writing overdue notices. Borrowers’ reactions appear positive: in a short survey, most participants said the messages were useful, although a few found them annoying. No borrower asked to leave the trial early.
Overall, the pilot seems worth repeating on a larger scale. A trial across all four branches, with a comparison group that does not receive messages, would show whether the effect holds for borrowers in general. The library board will discuss the proposal at its next meeting in September.
Find it
Tap the 7 hedges: every may, might and could, plus any form of suggest, appear and seem.
Check your understanding
1. Why might the effect be smaller across the whole library?
2. Which finding does the writer state with strong confidence?
3. What would a larger trial need in order to test the effect properly?
Sort it
Which control?
Tap a word, then tap the box it belongs in. Press Check when the tray is empty.
Unscramble
Calibrated claims
Tap the words in order to build the sentence. Tap a placed word to send it back.
1. One modal for possibility; no stacked hedges.
2. A reporting verb plus a limit on degree.
3. Name the evidence source: what people said about themselves.
4. Timing only, with no claim about cause.
5. Limit the claim to one group.
Crossword
Calibrate the claim
Click a clue or a square, then type. Press Check when you finish.
Across
Down
Choose the right amount of challenge
Communicate in three steps
The claim calibration panel. Pick one starting point, then move up only if the learner is ready. True or invented details are equally acceptable.
1
Personal answer
Make the language yours
Use the lesson’s earlier example or invent a similar situation. The details do not need to be personal.
Evaluate a fictional study and negotiate the strongest defensible public claim. Start with this cue: What was directly measured?
Next cues
Who and what conditions are covered?
Is the relationship causal or associative?
2
Guided role play
The claim calibration panel
Evaluate a fictional study and negotiate the strongest defensible public claim.
Role 1Researcher: present findings, limits and recommended wording.
Role 2Reviewer: challenge population, measurement, cause and certainty, then approve or revise each claim.
Useful language
What was directly measured?
Who and what conditions are covered?
Is the relationship causal or associative?
Where is strong wording justified?
3
Real-world challenge
Remove the support
Write 180 to 220 words about a fictional pilot study.
C1 target: Sustain a clear position for two minutes, adapt to a challenge, and reformulate one point more precisely.
Include
Limit population, time and measurement source.
Use one association claim, one cautious causal possibility and one justified booster.
Name at least two study limitations.
Write for a real reader
Write an evidence-calibrated summary
Write 180 to 220 words about a fictional pilot study.
Your text must
Limit population, time and measurement source.
Use one association claim, one cautious causal possibility and one justified booster.
Name at least two study limitations.
Reveal a possible model only after you have drafted your own version.
In a six-week pilot involving eighty employees, participants reported slightly lower stress and fewer missed deadlines. The findings suggest that flexible start times may have contributed to improvement for some teams. However, the pilot had no control group, and workload fell during the same period, so the schedule cannot be treated as the sole cause. Attendance records clearly show a 12 percent reduction in late arrivals. A larger controlled study is needed to determine whether this pattern continues across departments and over a longer period.
Reflect and transfer
Can you do this without the page?
Limit probability, scope, degree, cause and source precisely.
Use boosters when direct evidence warrants them.
Avoid empty hedge stacking and unsupported generalization.