Lesson 6: Science and risk
In this lesson we’ll look at how scientists investigate how much harm radiation does. This will help us begin to understand something of how science works.
We’ll look at how scientific discoveries are reported in the media and how people use this information to work out the risk to their own health.
It's unethical to do experiments on people that might seriously harm them
Scientists often do experiments but it’s wrong to do experiments on people that will harm them, such as exposing them to radiation.
So when scientists are investigating radiation and health they have to study people who have been exposed anyway. For example, because of the atomic bombs dropped on Hiroshima and Nagasaki or the Chernobyl nuclear accident. Scientists may also study people who are exposed to radiation as part of their job, like uranium miners and hospital workers.
The Chernobyl accident helped scientists understand more about radiation and health
We’re going to use an imaginary accident which is partly based on the world’s only major nuclear accident, at Chernobyl in 1986.
In this accident operators at a nuclear power station shut down some of the safety systems to conduct tests. They lost control of the reactor and there was a series of conventional (non-nuclear) explosions.
These spread radioactive fall-out over a large part of Europe and eventually over much of the Northern Hemisphere.
Science can only answer some types of question
There are some questions that science can answer and some that science can’t.
Science can’t answer questions about human values in the same way that you can’t use ideas about football to explain opera. Science CAN tell us whether something is possible, like building a nuclear reactor but it CAN’T tell us whether we should do it or not.
The land of scientific knowledge
Imagine science as being like exploring an unknown land.
Around the land is space that is unknowable by science. This space contains questions like: ‘Is there a God?’ ; ‘How do we get the most out of life?’; ‘Is research on animals wrong?’.
The scientific land that we want to explore contains answers to questions that can be observed or measured. ‘How old is the universe?’, ‘How can we reduce heart disease’, ‘How similar are rat and human organs?’.
How scientists develop theories
But scientists don’t just go round observing and measuring things and then pulling theories out of thin air.
Scientists have developed core theories that they treat as being ‘fact’ even though they know that later theories might replace them. One example would be the idea that everyday stuff is made up of particles which change their motion because of the forces between them.
Most science involves trying to describe precisely how one or more of these core theories apply to what is being studied. For example a chemist might make a new crystal that makes a very pure note when it’s tapped.
She might try and explain this by describing exactly what the particles are like, their separation and how they move. She might use this minor theory to make a prediction that she can test with an experiment. If the experiment doesn’t confirm the prediction she’ll assume there’s a problem with the details of her particular theory or experiment.
She won’t say that this crystal isn’t made of particles.
How theories change over time
Core theories can change over time. But scientists are very reluctant to take on a new one until they’re sure it’s better. The theory that the planets orbited the Sun not the Earth took 150 years to develop before it was much better at explaining the night sky.
To be accepted, a new core theory must explain everything the old core could plus much more.
What makes a good scientific theory?
A good scientific theory will suggest what to investigate next, though it takes the imagination of a scientist to work out the details. A core theory in biology is that DNA controls what our body does. This leads scientists to look for genes that might cause certain diseases.
Another sign of a good theory is that it predicts the existence of something that has never been seen. Scientists can then hunt for it.
For example the existence of the neutrino was predicted 20 years before it was finally detected.
Scientists increase scientific knowledge by using existing theories to gradually push out the boundaries of the known into the unknown. It takes time for the community of scientists to accept new research as ‘knowledge’.
The scientific community
So what is the ‘scientific community’ and how does it work to ‘do science’?
Most scientists work for universities, government research departments, or companies that use science to develop new products. A group of scientists may have an idea about a question they want to answer.
For example: “How has the nuclear accident affected the number of people who get brain cancer?”
They’ll need money to buy equipment and to pay themselves while they’re doing the research so they apply for a ‘research grant’.
Their application explains why the work is new and important and how it relates to what is already known. It explains what they want to find out, how they’re going to do it, how long they think it will take and what they will spend the money on.
This is essentially what you’re doing when you write a plan for your investigation. You just don’t get paid for it.
The university may have a fund of money to award as research grants or the scientists may apply to a government department, private company or maybe even a charity.
Let’s say our scientists get their money. How do they go about planning their research?
Getting valid data
Our scientists want to see whether the number of people with brain cancer after the accident is greater than before. This type of research is called an ‘epidemiological study’.
If their research shows an increase then it will be more convincing if it shows that more radiation exposure means a bigger risk of cancer.
It will be even more convincing if they can show that this agrees with other research showing HOW radiation causes brain cancer. So where are they going to get their data from? Hospital records, death certificates, newspapers?
Probably the best approach is to look at hospital records but how many do they need to see, all of them or just a sample? If it’s just a sample, how big does the sample have to be? You could do lots of random surveys and take an average but it probably makes more sense just to look at all the records.
So our scientists are going to look at all the tens of thousands of hospital records. But how many years do they need to look at? The problem is that the incidence of brain cancer changes every year. This is called scatter.
They have to use a branch of mathematics called statistics to see if they can spot any trends. One technique is to use a ‘moving average’ to flatten out the scatter.
Say our data showed that the rate suddenly increased in 1990, four years after the accident.
As well as removing scatter, scientists need to check that the incidence of cancer wasn’t changing anyway. The number of brain cancer cases might be changing for lots of different reasons. These are called confounding factors.
It might be that the number of cases is rising or it might just be that the hospital is getting better at detecting them. Perhaps a new technique for detecting brain cancer has been discovered, which would tend to show an increase in cases. Or maybe there was less money so fewer cases were being diagnosed, which would show up as a decrease.
Let’s assume that our scientists manage to adjust the data for all these effects and are sure that there HAS been an increase in brain cancer.
How do they know that this increase was caused by radiation from the nuclear accident?
Remember that the problem with cancer caused by radiation is that it looks just like any other cancer. The cancer isn’t radioactive and there aren’t any tests you can do to find out what caused it.
It would help if the scientists could show that an increase in radiation exposure increased the risk of getting brain cancer. This is much more difficult. The accident happened twenty years ago and all they have is some hospital records.
If someone shows up with brain cancer fifteen years after the accident you’ll know from the hospital records where they live now. But how do you know where they lived at the time of the accident? The only way to be sure is to follow up every brain cancer case.
If they do this then the scientists can plot a graph of relative risk of brain cancer against contamination. Here ‘relative’ means relative to your risk with no exposure to contamination. So with zero exposure your relative risk is 1. So the graph seems to confirm that increased radiation gave increased risk of brain cancer.
Publishing research in a scientific journal
So our scientists have done some research but nobody else knows about it.
Nobody has had a chance to look carefully at their work to check they haven’t made any mistakes. And other scientists can’t yet use it as a starting point for research of their own.
Our team will write up their research in a paper. It will be detailed and technical so that other scientists could try and duplicate their work if they wanted to check it.
There are around 10 000 scientific journals worldwide, most specialising in a tiny part of science. The scientists choose one that specialises in their area and send their paper to the editor. The editor will decide whether the work is original and important enough to merit publication.
Peer review: review by your equals
If she decides to accept it in principle then she sends copies of the paper to two or three working scientists, who act as referees. She will choose referees who specialise in the same area of science as the report’s authors. Only the editor knows the identity of all the referees. They don’t communicate with each other or the scientists who wrote the paper.
This use of referees is called ‘peer review’. Here ‘peer’ means ‘someone who is an equal’. So peer review means ‘review by your equals’ because science is not about who you are but the work that you do.
The referees may spend many weeks reading the paper to check it for mistakes or sometimes suggest improvements. Each referee then sends a brief recommendation to the editor whether to publish. She makes the final decision. If the editor decides not to publish she may suggest changes to be made to the paper if the scientists wanted to have another go.
Published work is there for the scientific community to use
So what does it mean if our scientists are successful in having their paper published in the journal?
Specialist scientific journals tend to be bought by universities and other institutions rather than individual scientists. Papers that have been published in peer-reviewed journals are considered to be part of the permanent scientific record.
But it doesn’t mean that their conclusions are ‘true’.
What it means is that the data and conclusions are now available for anyone to look at. Other scientists can draw different conclusions from the same data or they may use the conclusions to support their own.
Our research showed that brain cancer went up when radiation went up. This is called a ‘correlation’. But a correlation does not mean that one thing CAUSED another.
Our scientists concluded that radiation did cause the increase because other research showed a mechanism. Radiation damages DNA and damaged DNA can cause cancer.
But another scientist might have a theory that what causes brain cancer is fear. The people who were contaminated were afraid and this made them more likely to get brain cancer. Notice that this theory leads to predictions that can be tested. For example: people who worry less should be less at risk of brain cancer.
However our ‘fear causes cancer’ theory would not replace the ‘radiation causes cancer’ theory unless it could explain more things.
If a leading scientist was asked ‘Did the radiation from the accident cause brain cancer?’ then they might reply something like ‘Most research shows that it does and we know radiation damages DNA but some scientists have linked the increase to fear.’
Scientists try not to see peer-reviewed research that contradicts their own ideas as a threat but as an opportunity to learn more.
The people who pay the research grants may influence scientists' conclusions
It might be that the scientists feel under pressure to come to a conclusion that agrees with the group that funded them. That’s why some journals insist on knowing where the money came from. Was it the nuclear industry or perhaps an environmental group?
Even though science is about ideas, scientists themselves are still just people with beliefs and prejudices just like everyone else.
Open discussion of ideas is encouraged
So journals and conferences allow scientists to discuss their ideas both in writing and face to face.
How scientific research is reported in the media
So how do the public ever get to hear about scientific research? The simple answer is that they don’t, unless newspaper journalists think the research will make an interesting story.
Two articles can use the same figures but present very different points of view.
One article might the fact that 200 000 people may die. It doesn’t tell us that these are people who may die far in the future, not people who have died already in one great big group. It might show a graph that shows a big proportional increase in certain type of cancer but it doesn’t say that the risk is still very small. It wants to convince us that nuclear accidents are very bad because they affect a large number of people.
Another article might the fact that the overall cancer rate has increased by less than 1%. It doesn’t tell us that this still amounts to a lot of people in a big population. It might show a graph that shows that the risk of all cancers has stayed pretty much the same but it doesn’t tell us about the human effects of each extra case. It wants to convince us that nuclear accidents are not so bad because your risk of getting cancer hasn’t increased very much.
So if you were asked ‘How bad was the nuclear accident?’ your answer may depend on which article you happened to read.
Nothing is 100% safe
Both articles would acknowledge that the research shows about 3000 extra people a year would die of cancer. Can this possibly be justified? One way to consider this figure is in relation to other risks that people take.
Over 100 000 people die each year in the UK because of heart disease. If the government banned smoking, high fat food and extra salt and made everyone take regular exercise this figure could be reduced to almost zero!
What would you want to see prohibited to save lives? Shouldn't people be able to take risks if they want to?
Almost everything we do carries some risk. Nothing is ever ‘safe’.
It's easy to spend lots of money without decreasing risk much
It also costs money to make things safe.
Say it costs $1 million to make a steel surround for a nuclear reactor that will stop a leak for 90% of all possible accidents. It might cost another $1 million to make it stop a leak for 95% of all possible accidents. The next 5% of protection costs as much as the first 90%.
The ALARA principle
So people and governments have to work out what level of risk is As Low As Reasonably Achievable. This means you don’t waste millions reducing one risk a tiny amount when you could spend it reducing another risk by a lot.
The Precautionary Principle
Of course one way to reduce a risk to zero is to avoid it altogether. For example you don’t need to go parachuting. Perhaps we should avoid doing things where the result might be serious and irreversible harm.
This is called the ‘precautionary principle’ and was first widely adopted by governments in the 1980s.
It makes sense to avoid doing something if you don't know whether serious and irreversible harm might result from it.
An example might be ‘We don’t know the long-term risks of burying nuclear waste so we should avoid doing it.’
One problem with the precautionary principle is that it’s never possible to prove that something won’t cause harm. Another is that it ignores both the potential benefits and the cost of doing nothing. The precautionary principle seems simple but people disagree on the meanings of ‘know’, ‘serious’, ‘might’ and ‘harm’.
Do the benefits outweigh the costs?
A more common way of working out whether or how to adopt a new technology is to add up the costs and benefits. A cost of nuclear power is the risks posed by nuclear waste. A benefit is the electricity produced.
People accept a risk if they think it is outweighed by the benefits. For example if you value electricity more than the risk of nuclear waste.
But how do we know how big a risk is?
The worst type of risk is likely to happen and the consequences are very bad. Getting lung cancer from smoking would be a good example.
There has only ever been one nuclear accident, Chernobyl, where people outside a plant died. How would you rate the risk of a nuclear power station accident?
Why people overestimate risk
Nuclear radiation is not a big risk factor for cancer compared with smoking, poor diet and lack of exercise. Nuclear accidents are extremely rare but still many people rate nuclear energy as being extremely risky.
Why is this? A possible reason is that when people think about taking a risk they don’t just think about numbers. In fact the actual numerical size of the risk may not feature at all in how people rate it.
For example people rate risks that are forced on them higher than risks they decide to take by choice.
Science can’t answer questions about whether there should be nuclear power. It can only provide the numbers. It is up to each person to decide the risks they are prepared to take themselves or pass on to other people.