Look at the numbers
Science relies heavily on statistics, and the bigger the numbers, the more accurate they are likely to be. As a general rule, the larger the study, the more reliable it is.
Let’s take the fictional example of a group of five people with cancer. Asking about their diet, we find out that four of them eat sausages for breakfast every day.
This suggests that four out of five people who get cancer breakfast on sausages, which could be interpreted to mean that eating sausages is very strongly linked to developing cancer. That sounds rather high - and certainly enough to put you off a cooked breakfast.
But what if we take a group of 5,000 people with cancer, and find that only 200 of them ate sausages daily, rather than the 4,000 that we might predict from our smaller study.
This tells us that it’s likely the results of the first study were a statistical “blip,” and sausages have a much smaller impact on cancer risk. In this case, the larger study would be a more accurate reflection of the situation in real life.
However, neither of these imaginary studies tell us for sure whether eating sausages for breakfast actually causes cancer.
Why?
Because we don't know about the eating habits of people who don't get cancer. Perhaps they eat even more sausages, which might suggest that sausages can prevent cancer! So we need a control group.
Read our other Top Tips:
Increase text
Decrease text
Print this page
