What Is a Survey? Definition and Examples
What is a survey? What is a census? What is the difference between a survey and a census? A survey is the collection of information from all members of either a sample or a population.
In this lesson, you will learn
- what a survey is, and the difference between a sample survey and a census
- why a census is often hard and expensive, and why a census does not need a large population
- how a sample survey gives an estimate, and why different samples give different results
Sample Survey and Census: Two Kinds of Surveys
If the information is collected from all members of the sample, the survey is called a sample survey.
If the information is collected from all members of the population, the survey is called a census.
Notice that a census may not be easy to conduct since the size of the population is usually quite large.
It could also be very expensive if you are paying people to take these surveys.
The United States of America tries to obtain a census every 10 years. It fails every time since it is not possible to reach everyone.
Keep in mind that to call a survey a census, the population does not have to be quite large.
For example, suppose you conduct a survey on incomes of those who graduated from Harvard University in 2000.
Suppose that 1,500 students graduated from Harvard in 2000.
A census may be conducted here as long as the university kept the contact information of all 1,500 students who graduated in 2000. Furthermore, the university was able to get an answer from every single graduate.
If the university does not have the contact information of all 1,500 students who graduated in 2000, they may have to conduct a sample survey instead of a census.
A teacher asks all 24 students in her class which day they want the field trip. Is this a census or a sample survey?
Show answer
A census. The population is the 24 students in the class, and every one of them is asked. A census can be small.
Census or Sample Survey: Which One to Use?
Both kinds of surveys collect information, but they give different kinds of answers. Here is how they compare.
| Census | Sample survey | |
|---|---|---|
| Who is asked? | Every member of the population | Only the members of a sample |
| Cost and time | Often expensive and slow for a large population | Cheaper and faster |
| Result | The exact value for the population, if everyone answers | An estimate of the value for the population |
| Good choice when | The population is small, or you can reach everyone | The population is large or hard to reach |
- The population is all 1,200 students. The sample is the 60 students who are asked.
- In the sample, 18 out of 60 walk: 18 ÷ 60 = 0.3, or 30%.
- Use the sample to estimate for the population: 30% of 1,200 = 0.3 × 1,200 = 360.
About 360 students walk to school. Only a census, asking all 1,200 students, would give the exact number.
Thinking that any survey with many answers is a good sample survey. If people volunteer themselves to answer, for example on a website, the survey is self-selected. People with strong opinions are more likely to answer, so the sample may not be a representative sample of the population. Choosing the people at random is better.
A city of 80,000 people wants to know what its residents think about a new bus line. Should it use a census or a sample survey?
Show answer
A sample survey. Asking all 80,000 residents would be expensive and slow. A random sample of residents gives a good estimate at a much lower cost.
Interactive: Run a Sample Survey, Then a Census
A school has 400 students. How many of them walk to school? Ask random samples of different sizes, then take a census to see the exact answer. Asking one student costs $2.
walks to school does not walk not asked
Sample size
Try several samples of 10 and several samples of 100. Small samples usually give results that are farther from the census answer. To learn more, see sampling and nonsampling errors and common sampling methods.
Summary
- A survey is the collection of information from all members of either a sample or a population.
- A survey of all members of a sample is a sample survey. A survey of all members of the population is a census.
- A census may be hard and expensive when the population is large. The United States tries to take a census every 10 years.
- A census does not need a large population: asking every one of 1,500 graduates is also a census.
- A sample survey is cheaper and faster, but it gives an estimate. Different samples give different results, and the sample should be chosen at random.
Test Yourself: Interactive 10-Question Quiz
Now check what you learned. Every answer is explained, and each time you play you get new questions.
Quiz complete!