A sample can be unfair — Data, 11–13 years
A sample may be easy to collect but still fail to represent the larger group. Looking at who was included, who was missed and how people were chosen helps reveal sampling bias.
A sample should mirror the group
Usually we study a smaller sample to learn about a larger group. A useful sample has the important kinds of people, places or objects in roughly the right balance. It does not need to copy every detail, but it should not leave one part out just because it was harder to reach.
Why does selection matter?
The problem is that the easiest people to ask may not be like everyone else. If you ask only the players on a team whether the school needs more sports, the answers will lean in one direction. The result may describe the people asked, not the whole school.
A fairer school survey
A school wants to know how students travel there. Asking the first 20 students at the bike racks will overrepresent cyclists. Instead, choose students from every year group and several entrances, using a method that gives each student a chance to be selected. Then compare the answers with the school’s full range of students.
Large is not automatically fair
It is reasonable to think that asking more people must improve the answer. But 1,000 answers from one narrow group can be less useful than 100 answers chosen from the whole population. Size reduces some random wobble; it does not repair a biased choice of people.
Surveys and decisions
Governments, companies and researchers use samples when asking everyone would cost too much time or money. A food survey, an opinion poll or a product test can guide a real decision. Before trusting the result, check who was invited and whether some voices were difficult to include.
Keep exploring
Other languages
Loading MyLeoNes™…