The vocabulary of a study
Eight terms. Match each to what it means.
The same research, four times
Four texts about one study of weekly food-waste collections. Read for what each one claims, then read again for what has been removed between the first and the second.
A · From the abstract
We report results from a five-year study of 1,400 households in two mid-sized European cities, following the introduction of a weekly food-waste collection. Participating households reduced measured food waste by a mean of 18% (95% CI 11–24%) relative to controls over the study period. The effect was largest in the first eighteen months and appeared to attenuate thereafter. Reductions were not evenly distributed: households with children under five showed no statistically significant change. Both cities had existing kerbside recycling, and the finding may not generalise to areas without it. Further work is needed to establish whether the observed attenuation reflects habituation or measurement drift.
B · The university press release
New research proves that weekly food-waste collections cut household food waste by nearly a fifth.
The five-year study, the largest of its kind, shows that simply changing collection schedules can deliver dramatic environmental benefits at almost no cost. “This is a game-changer for local authorities,” said the lead author. Councils across the country are now expected to follow suit.
The findings come as households face growing pressure to reduce their environmental impact, and demonstrate that small changes in policy can make an enormous difference.
C · From a newspaper feature
The study is a good one and the press release is not. That gap is not unusual: research on exaggeration in science reporting has repeatedly found that most of the overstatement in a news story is already present in the university’s own release, which is worth remembering the next time a journalist is blamed for it.
What the paper reports is an 18% reduction, with an interval running from 11 to 24, in two cities that already had kerbside recycling, with the effect concentrated in the first eighteen months and absent altogether in households with young children. What the release reports is that the research “proves” that collections “cut” waste by “nearly a fifth” — with no interval, no scope, no attenuation and no exceptions.
Every one of those omissions is a hedge that has been removed rather than a fact that has been added. That is the specific mechanism of this kind of inflation, and it is why it is so hard to point at: nothing in the release is false.
D · From a comment thread
so basically councils are wasting our money on bin collections that don’t work
This proves what I’ve been saying for years, people just need to buy less food
Everyone knows nobody actually uses the food waste bin
It always works in Germany so obviously it works everywhere
I mean 18% is nothing, that’s basically a rounding error
typical, one study and suddenly it’s settled science
Eight statements. Decide which text each is true of. A text may be the answer more than once.
A The abstract
B The press release
C The feature
D The comment thread
1. States a condition under which the finding may not hold.
2. Uses the word proves about a single study.
3. Points out that nothing in the inflated version is actually false.
4. Generalises from one country to all countries in a single sentence.
5. Reports a group in which no significant effect was found.
6. Predicts the behaviour of organisations that were not studied.
7. Locates the source of the overstatement before the journalism.
8. Dismisses a measured effect by calling it a rounding error.
Count what was removed. Between the abstract and the release, five hedges disappear: the interval (11–24), the scope (two cities with existing recycling), the attenuation (largest in the first eighteen months), the exception (households with young children) and the source-limit (one study). Nothing false was added.
That is the shape of nearly all over-claiming, in press releases and in student essays alike: not invention, deletion. Which means the repair is not fact-checking — it is putting the slots back.
A fifth text, with four sentences missing
Four sentences have been removed. Put each one back where it belongs.
The reusable cup is the most photographed object in environmental behaviour and one of the least studied. [1]
A ceramic mug has to be used somewhere between twenty and a thousand times before it beats a paper cup, depending on how it is washed, how the electricity is generated, and what you decide to count. [2] The range is that wide because the studies are measuring different things in different countries under different assumptions.
[3] It means that the honest sentence is longer than the sentence anybody wants: reusable cups appear to be better in most circumstances, provided they are actually reused, and the number of reuses required varies by roughly two orders of magnitude.
None of which makes the cup pointless. [4]
It makes it a small, conditional good that has been asked to carry an argument a great deal larger than itself.
Putting the slots back
Eight gaps, from a summary written for a council committee. Choose the option that claims exactly what the study supports.
A five-year study of 1,400 households that weekly food-waste collections are associated with lower measured waste. The reduction was eighteen per cent, with an interval from eleven to twenty-four. The effect to have been largest in the first eighteen months. It was observed in two mid-sized cities that had kerbside recycling, so it generalise to areas without it. No significant change was found among households with children under five, which the authors explain. On this evidence a trial be justified here; a full rollout premature.
Eight sentences. Seven claim more or less than the evidence supports; one is correct as written.
1. This study proves that food-waste collections work.
2. The scheme reduced waste by 18%. — the interval ran from 11 to 24
3. It works everywhere. — two cities, both with existing kerbside recycling
4. Families with small children do not respond to these schemes. — no significant change was found in that group
5. I think that maybe it could possibly be the case that the scheme may perhaps help.
6. The effect attenuated after eighteen months, which proves people lose interest.
7. The finding may not generalise to areas without kerbside recycling. — correct as written
8. Everyone knows these schemes are a waste of money.
Write the press release back
Task card · Writing and speaking · Pairs · 12 minutes
Text B removed five hedges from text A. Your job is to put them back without making the release unpublishable — because a press office that says nothing gets no coverage, and a press office that over-claims gets the wrong coverage.
Sixty words, honest, and still worth printing.
- 3 minutes: list the five things text B removed. Interval, scope, attenuation, exception, source.
- 6 minutes: rewrite the opening two sentences of the release so that all five are present, and it still reads like a press release rather than an abstract.
- 3 minutes: read yours aloud against the original. Ask one question of each version: what would a councillor do differently after reading this?
- The test is not which is more cautious. It is which one a reader could act on without being misled.
The hard part is that the honest version is longer. Every slot you fill costs words, and a press release has very few. That is a real constraint and not an excuse: the question is which hedges are load-bearing. Here, scope and the exception change what a council should do; the confidence interval mostly does not.
Which hedges were load-bearing
Two questions.
1. Which removed hedge mattered most to a reader deciding what to do?
2. Could you keep all five and still sound like a press release?
Deletion, not invention
Two questions. Twenty seconds.
1. How does over-claiming usually work?
2. What does “no significant change was found” mean?
Next. The instruments themselves, sorted — modal, lexical and adverbial — and the question of which slot each one can actually reach.