When you get a blood test, does the nurse take out all of your blood and look at it? (All of your blood would be the population of blood.) Obviously, they take a sample. Do you have to drink all of the milk in a gallon to see if it is spoiled? Once again, it is logical to just take a small sample to get the information that you need.

If you are doing a project and don’t want to work with the entire population, you too can use a sample of the population. While it is true that anyone who is part of the population can be asked to participate, the only way your project will get accurate information is if you randomly select people to be involved. If you want to know the favorite sport in school and your sample all comes from the wrestling team, will the answer to your question really reflect the favorite sport of the population?

A random sample means that everyone in the population has an equally likely chance of being selected. Let’s pretend that you want to compare the number of hours each grade works on homework. Here are some ways you might get a random sample.

• Go to the cafeteria and ask the person at the left corner seat of every table their grade and the average number of hours they do homework each day.

• Give every student a number. Write the numbers on a paper and randomly select the numbers of students to participate. You could also use a random number table or a graphic calculator or computer to generate a list of random numbers

What other ways might you get a random sample?

The most difficult question is how many should be included in the sample? There is a method based on the Central Limit Theory that can tell you about how big your sample should be, but for our purposes, we will say that you don’t want your sample to be too small. If you have a question about sample size, you may want to ask your teacher to help you determine the number of people you should use in your project.

There is generally some error if you are conducting an experiment or a survey. This is called sampling error and the amount of error is generally told to you when you read about the project. The bigger your sample, the less error you should expect. For example, a survey might say that 55 percent of the people in the U.S. have football as their favorite sport (with a sampling error of ±3 percent. This means that the real number is between 52 percent (55 - 3) and 58 percent (55 + 3).