Skip to content

Hedging: saying exactly how sure you are

Verifiedintermediate6 min

The vocabulary of scientific caution, and why over-claiming costs more than it gains.

Judges and reviewers read claim strength closely. Matching your language to your evidence is a skill worth practising deliberately.

Over-claiming, and the fix:

  • "This proves that…" → "These results are consistent with…" (Experiments support; they do not prove.)
  • "Fertiliser X is the best." → "Fertiliser X produced the greatest mean biomass of the three tested."
  • "There was no effect." → "No significant difference was detected (p = 0.31); with n = 5 per group this study had limited power to detect small effects."
  • "This will help farmers." → "If the effect holds at field scale, this approach may be relevant to…"

Correlation language. If you did not manipulate the variable, do not use causal verbs. "Was associated with", not "caused".

Own your limitations. A Discussion that names three real limitations reads as more competent, not less. The student who says "my sample size was too small to detect anything under a 20% difference" is demonstrating exactly the judgement judges are looking for.

Precision is not timidity. "We found a 34% increase (95% CI 18–50%)" is a stronger sentence than "we found a huge increase", because it can be checked.

More in Scientific Writing

Verifiedbeginner

The IMRaD structure

What belongs in Introduction, Methods, Results, and Discussion — and the sentences that end up in the wrong one.

  • imrad
  • structure
  • paper