Health and drugs · Investigation
Substance misuse and decisions
Nobody decides anything from a table of facts. They decide from what they have heard, from what they think everyone else is doing, and from claims that sound like evidence and are not. This lesson is about telling the difference.
Start here
“It's natural, so it can't do you much harm.”
Deadly nightshade is natural. Foxglove is natural. Tobacco is a plant, and so is the source of the strongest painkillers in the hospital. Every one of them grew out of the ground without human help, and the ground has no opinion about your health.
Which of these actually tells you whether a substance will harm you?
Only the last one. Where a molecule came from, what a shop is allowed to sell, and how many people you know who use it are all facts about the world around the substance. What it does to a body, at that amount, is the only question with a biological answer — and answering it takes evidence of a particular shape.
At the bench · five claims, five faults
Find the fault in the evidence
0 of 5 checked
Each claim comes with the evidence someone offered for it. Every fault in the list below is the real fault of one of these five claims, so a wrong answer still teaches you something about the claim you picked it for.
The claim
“Cannabis is safe — it is a plant, and people have used it for thousands of years.”
“Most people our age vape — it is normal now.”
“Energy drinks improve your exam results.”
“This herbal supplement cures anxiety — nine out of ten users agree.”
“Drinking every day cannot be that harmful — my grandad did it and lived to 90.”
The evidence offeredIt grows in the ground with no chemicals added, and it has been used in many cultures for centuries.
The evidence offeredTwenty-two of the thirty people in my friendship group and my form have tried it, and I see people vaping outside school every day.
The evidence offeredA study of 500 sixth-formers found that those who drank energy drinks during study leave scored on average four marks higher than those who did not.
The evidence offeredReviews on the seller’s own website: 90% of the people they asked said they felt calmer after taking it for a month.
The evidence offeredOne person, known personally, who drank daily for sixty years and stayed well.
What is wrong with that evidence?
That is the faultNot the main fault here
Where the substance came from is not evidence about what it does.
Both halves of the evidence are about the history of the plant rather than its effect on a body. Deadly nightshade also grows in the ground and has also been known for centuries. Long use tells you a substance does not kill most people quickly, which is a much weaker claim than safe — the harms being argued about here are effects on the developing brain that appear over years.
What would settle it: Studies following large numbers of users and non-users forward over years, recording mental and physical health outcomes, with the users’ use recorded before anything went wrong.
That is the faultNot the main fault here
The people asked are not a fair sample of the group being described.
Your friendship group is not a random sample of your age, and neither is the group standing outside school — people who vape are visible and talk about it, people who do not are doing nothing you would notice. This is the same fault as testing a new drug only on people who volunteered because they already liked it.
What would settle it: A national survey that samples tens of thousands of pupils chosen at random rather than by who is easy to see, and asks about regular use rather than about ever trying it once.
That is the faultNot the main fault here
Two things happening together has been treated as one causing the other.
The two groups chose themselves, and they differ in more than the drink. Students who revise for long hours late at night are exactly the students who buy energy drinks, so revision hours could produce the whole difference. The study is large and still cannot tell you which way the arrow points — size fixes noise, not confusion.
What would settle it: Take one group of students, split them at random into two, give one the drink and the other an identical-tasting drink with no caffeine, and compare. Randomising is what breaks the link between the drink and everything else about the person.
That is the faultNot the main fault here
There is no comparison group, and the people asked were chosen by the seller.
Nobody was compared with anything. Anxiety varies week to week and improves for many people anyway, so “felt calmer after a month” would be reported by a good fraction of people given a tablet with nothing in it. On top of that, the seller chose which reviews to publish and profits from the answer, and no dose is stated at all.
What would settle it: A trial where neither the patients nor the assessors know who received the supplement and who received an identical dummy, with the results published by someone with nothing to sell.
That is the faultNot the main fault here
One person’s story cannot tell you the risk for anyone else.
Risk is a statement about how often something happens across many people, so a single case can never contradict it — people survive things that are dangerous all the time. There is also a hidden bias in who you get to hear about: the drinkers who died at 55 are not around to be anyone’s example, so the survivors are the only ones telling you their story.
What would settle it: Death and illness rates for many thousands of daily drinkers compared with similar non-drinkers, which is how the risk was estimated in the first place.
Four questions, asked in this order
You do not need to know the science to test the evidence.
How many, and who?
Thirty friends is not a sample of a nation. Ask how the people were chosen, not just how many there were.
Compared with what?
A result with nothing to compare it against is not a result. Who was the group that did not get the thing?
Who is telling me?
Someone who profits from the answer is not disqualified, but their evidence needs checking by someone who does not.
Together, or because?
Two things rising at once is where investigations start, not where they finish. Something else may be causing both.
Key fact
A claim about a drug is only as good as the evidence behind it. Ask how many people were studied, whether there was a group to compare against, who paid for the answer, and whether two things happening together has been mistaken for one causing the other.
Think again
“If it's natural, it's safe — it's the chemicals that hurt you.”
Every substance in this lesson is a chemical, including the water you drank at break. A plant does not manufacture its molecules for your benefit; several of them are made specifically to poison whatever tries to eat the plant, which is why nicotine exists at all. Ricin, cyanide and the digitalis in foxglove leaves are natural, and the purest paracetamol in the pharmacy is made in a factory to a known dose, which is precisely what makes it safe to use. So natural and synthetic tell you about a molecule's history and nothing about its effects; the questions that matter are what it does in the body and in what amount. Herbal products carry an extra problem the pharmacy does not: the dose in a leaf varies with the plant, the season and the soil, so even a genuinely useful herbal remedy is a medicine of unknown strength.
“Everyone my age is doing it.”
Ask a year group to estimate what fraction of them drinks, smokes or vapes regularly, and the average guess comes out far above the real figure — every time this is measured, in every country it is measured in. The reason is a sampling fault you have already met: the people doing it are visible and talk about it, the people not doing it are not doing anything noticeable, so your sample is the loud end of the room. National surveys of tens of thousands of pupils consistently put regular use in a clear minority at this age. This is not a comfort exercise. The misjudged figure does real work, because believing something is normal is one of the strongest predictors of starting it — which means an inaccurate estimate is a genuine cause, and correcting it changes behaviour on its own.
Mastery ladder
Not started yet.
Rungs 3 and 4 you mark yourself.
Rung 1 · Name the fault
An advert says a sleep supplement works because 94% of the customers who reviewed it on the company’s website slept better. What is the main fault?
Rung 2 · The one that catches people
Someone answers a warning about daily drinking with “my grandad drank every day and lived to 90”. Why does that not weaken the warning?
Rung 3 · Design the test
A company claims its drink improves reaction times. Design a fair test of that claim, and say why each feature of your design is there.
Rung 4 · Take it somewhere new
A video shared in a group chat has someone in a white coat saying a substance is “completely safe, and the studies prove it”. It has two million views and thousands of agreeing comments. Work out what you actually know from this, and list what you would check before believing it.
Key note
Substance misuse means using a substance in a way that damages health — including legal ones, and including medicines taken wrongly. Claims about drugs are tested the same way as any other claim: how many people, compared with whom, reported by whom, and does the evidence show a cause or only two things happening together. Where a substance came from tells you nothing, and what you think everyone else is doing is usually wrong.
Going further
In 1950 the link between smoking and lung cancer was an argument, not a fact — plenty of doctors smoked, and tobacco companies pointed out, correctly, that a correlation is not a cause. Richard Doll and Austin Bradford Hill answered it properly: they recruited around 40,000 British doctors, recorded what each one smoked, and then waited. Following the same people forward for decades meant the smoking was recorded before anyone became ill, so it could not be explained away as sick people misremembering. The death rate climbed with the number of cigarettes, and it fell in the doctors who stopped. That study design — large, prospective, dose-related, reversible on stopping — is why the link is now considered settled, and it is the standard every claim in this lesson is measured against.
If any of this is about you or someone you know
This lesson teaches how to test a claim, which is the science. What to do about pressure from people you like, and what the law says, belong in your PSHE and RSE lessons and are worth taking there. If any of this is about your own home, your own use, or someone you are worried about, talk to a trusted adult, the school nurse, a pharmacist or a GP — and a pharmacist will answer a question about a substance without an appointment and without a lecture. Your school's PSHE materials list the national confidential services by name.
Before this lesson
Connects to
At GCSE this becomes
- Evaluating evidence: sample size, control groups, placebo and double-blind trials, peer review, and the difference between correlation and cause.
Where to next
Ask Mr Badmus AI
Want to test a claim you have actually seen?
If any of this is about you or someone you know, talk to someone you trust — a parent or carer, a teacher, your school nurse, a pharmacist or your GP. Out of school hours: Childline — 0800 1111, free and confidential. FRANK — 0300 123 6600.
The five claims are written for this lesson. They are the shapes of arguments people really make, not quotations from named sources, and the studies described in them are invented examples with invented numbers. The doctors' study in Going further is real. Nothing on this page is medical or legal advice.
Lesson content © MrBadmusAI.