Scientific Thinking » How Science Works

Scientific ideas change over time

A scientific theory is the best explanation we have for the evidence we have so far. When new data or new equipment arrives, a theory may be changed or replaced. This is not a failure. It is how science improves.

Here is an example from medicine. For hundreds of years, doctors followed the ideas of the Greek physician Galen, who taught that blood was made in the liver and then used up by the body. In 1628 William Harvey measured how much blood the heart pushes out and showed that the body could not possibly make that much new blood each hour. His data supported a different idea: blood must be circulating around the body again and again.

A second example is stomach ulcers. For most of the 20th century, doctors believed ulcers were caused by stress and spicy food. In the 1980s, Barry Marshall and Robin Warren found a type of bacterium in the stomachs of patients with ulcers. After more evidence, including successful treatment with antibiotics, the old explanation was replaced.

Key terms:

• Theory: an explanation of observations that is supported by a large amount of evidence.
• Evidence: data from experiments or observations that is used to support or reject an idea.

Does the data support the theory?

You may be asked to decide whether data supports a theory. Follow three steps:

1. State what the theory predicts.
2. Compare that prediction with the data.
3. Say whether it fits, and whether the data is enough to be sure.

Data that fits a theory does not prove it is true for ever. It just makes the theory more likely. Data that clearly does not fit is a reason to change the theory.

Models in science

A model is a simplified way of describing or showing something that is hard to see directly. Models are used to explain ideas, to make predictions and to solve problems. A model can be tested by comparing what it predicts with real observations or experiments.

Every model has limitations. A diagram of the heart is flat and still, but a real heart is 3D and moving. A good scientist knows what a model shows well and where it stops working.

Key terms:

• Model: a simplified representation of a real object, idea or process.
• Prediction: a statement of what you expect to happen, based on a model or theory.

The power and limits of science

Science is powerful. It has given us vaccines, heart surgery and ways to feed a growing population. But it cannot answer every question. Some questions cannot be answered because:

• the data is incomplete, because we have not yet found out enough
• the data is not available, because we cannot yet measure it or the event happened in the past
• the answer is uncertain, because results vary and there is more than one possible explanation

Science also cannot decide questions of opinion, such as whether something is right or wrong. Science can tell us what we can do. It cannot tell us what we should do.

Using science: benefits, costs and ethics

New technologies often help people, but they can also bring problems. When we judge a new use of science, we look at four types of effect:

An ethical issue is a question about right and wrong. For example, should new medicines be tested on animals before they are given to people? One argument for says that it protects patients from harmful drugs. One argument against says that animals can suffer. Different people weigh up the arguments differently, which is why such issues need discussion, not just data.

To make a good decision you should look at the evidence, compare the benefits with the harms, and listen to the arguments on each side.

Risk

A hazard is something that could cause harm. A risk is the chance that the harm will actually happen. Almost everything has some risk, so scientists ask how big it is and whether the benefits are worth it. A new treatment might have a rare side effect, but still be used because it saves many lives.

Risk is measured with data, for example the number of people harmed out of every 100,000. But people often perceive risk very differently from the measured risk. The spec gives these reasons:

• Voluntary or imposed: we accept risks we choose, such as playing contact sport, more easily than ones forced on us.
• Familiar or unfamiliar: everyday risks seem smaller than new, strange ones.
• Visible or invisible: a hazard we can see feels more dangerous than one we cannot, such as a gas with no smell.

Peer review and communicating results

Scientists share their work with other scientists and with the public. Before results are published in a scientific journal, other experts check them. This is peer review. It helps to spot false claims and mistakes, and it helps scientists to agree which claims should be accepted as valid.

Results are also reported in newspapers, on websites and on social media. These reports are not peer reviewed, so they may be oversimplified, inaccurate or biased. Scientists need to communicate to different audiences, from other scientists to schoolchildren, and each needs a different style.

When you read a news story about science, ask: who did the research, how many people or samples were used, and has it been checked by others?

Exam-style question

A newspaper headline says: "Eating blueberries cures memory loss". The story is based on one study of 12 people which has not been peer reviewed.

(a) Suggest why the headline might not be reliable. [2 marks]

(b) Explain how peer review would help. [2 marks]

(c) Give one reason why people may see a hazard as more dangerous than it really is. [1 mark]