Choose who to ask

If a school wants to learn which lunch students like, it could ask students from different grades. Asking only one class might miss what other students think. How a sample is chosen can affect what the results show. 1 2

Use a sample with care

A sample gives clues, but it may not match every person in the larger group. A survey of ten students cannot prove what all students prefer. Reports should say who was asked and how many people answered. 1 2

A larger sample can still be biased

Asking many people does not fix a survey that excludes some groups. If only students who ride the bus answer a travel survey, it misses walkers and car riders. A sample should represent the group the survey is meant to describe. 1 2