# What is a representative sample in sociology?

> Written with AI · [Sociology: research methods](https://lessoons.com/en/sociological-research-methods) · Grade 11 · 2026-11-12 · https://lessoons.com/en/sociological-research-methods/what-is-a-representative-sample-in-sociology

A representative sample is a small group of people chosen so that it has the same makeup as the whole population being studied: the same mix of ages, places, incomes and other features that matter. Sociologists use one so that what they find in the sample can fairly be said of everyone. Random and stratified sampling aim at it; a convenience sample usually misses it.

## The topic explained

### Population, sample and sampling frame

Sociologists rarely ask everyone. The population is everyone a study is about: all 11th graders in Israel, all the residents of a town, all Israelis aged 20 and over. A sample is the part of the population that is actually studied. It is representative when it has the same makeup as the population in every feature that matters to the question, so that the findings can be generalized, carried over from the sample to the population.

To pick a sample, researchers need a sampling frame, a list of the population from which names are drawn: a school's register, a list of addresses. A sample can be no better than its frame. Israel's Central Bureau of Statistics has run its Social Survey every year since 2002 on a new sample of about 7,000 people aged 20 and over, chosen so that it stands for the whole adult population. The misreading to avoid: representative does not mean large. It means that the sample is a small copy of the population.

### Three ways to choose a sample

In a simple random sample every member of the population has the same chance of being chosen, as in a lottery: numbers are drawn from the frame by a computer. Chance does the work of fairness, though a small random sample may by luck include too few members of a small group.

A stratified sample prevents that. The population is first divided into strata, groups such as grades, districts or age groups, and each stratum is sampled in proportion to its share of the population. A convenience sample takes whoever is easy to reach: the friends of the researcher, the people in a mall at noon, the readers who answer an online poll. It is cheap and fast and useful for trying out a questionnaire, but it cannot be called representative.

### A worked example: a stratified sample

A made-up example: a school of 1,000 students has 400 in 10th grade, 350 in 11th and 250 in 12th, and a student researcher wants a sample of 80. Each grade gets its share. 10th grade is 400 ÷ 1,000 = 40% of the school, so it gets 0.40 × 80 = 32 students; 11th grade gets 0.35 × 80 = 28; 12th grade gets 0.25 × 80 = 20. The check: 32 + 28 + 20 = 80. Within each grade the students are then drawn at random from the register. Why bother? A simple random sample of 80 would give each grade its share only on average: by chance it might take 30 students from 12th grade and only 20 from 10th. Stratifying guarantees that every grade appears in its true proportion, which matters when the grades may answer differently.

### How a famous poll went wrong

In 1936 the American magazine The Literary Digest polled about 10 million people on the presidential election, and 2.38 million answered. It forecast that Alf Landon would win 57% of the vote. Franklin Roosevelt won 60.8% and every state but two, and Landon about 36.5%. George Gallup, with a sample of about 50,000, called the winner correctly.

Two biases had sunk the giant poll. The lists came from the magazine's readers, car owners and telephone users, who in 1936 were better off than most voters: a sampling bias in the frame. And only about a quarter answered, and those who answered differed from those who did not, which later research found was the main cause: a non-response bias. The lesson for every bagrut answer is that a biased sample stays biased however large it grows.

## Practice: representative samples

For 11th grade students of sociology. The school and youth movement figures are made up for practice. Answer from what is printed, then check your work against the answers.

1. **A researcher wants to know how Israeli high-school students spend their evenings. She questions 600 students from 20 schools. What is the population, and what is the sample?** — The population is all high-school students in Israel. The sample is the 600 students who were actually questioned.
2. **Write the missing step: from what does a researcher draw the sample?** — A sampling frame: a list of the members of the population, such as a school register, from which the sample is drawn.
3. **Simple random, stratified or convenience? (a) a student asks the first 30 people who leave the cinema; (b) a computer draws 200 names at random from the list of all members of a club; (c) a researcher divides a city into its neighborhoods and samples each one in proportion to its population.** — (a) convenience: whoever is at hand; (b) simple random: each member has an equal chance; (c) stratified: the neighborhoods are the strata, each sampled in proportion.
4. **A youth movement has 600 members in three branches and wants a stratified sample of 50. Fill in the number from each branch.** — North: 50 × 0.50 = 25. Center: 50 × 0.30 = 15. South: 50 × 0.20 = 10. Check: 25 + 15 + 10 = 50.
5. **Read the chart. By how many percentage points did the Literary Digest's forecast for Landon miss his result?** — By 57 - 36.5 = 20.5 percentage points. The poll was huge, 2.38 million answers, and still wrong, because the sample was biased.
6. **A news website asks its readers: “Do you think teenagers spend too much time on screens?” 40,000 readers vote. Why is the result not representative of Israeli adults?** — It is a self-selected convenience sample: only readers of that site took part, and only those who chose to vote, who may care more about the subject than others. Its size does not fix the bias.
7. **Plan a representative sample of 12th graders in Israel for a study of plans after school. What strata would you use, and why?** — A good answer proposes strata that may affect plans after school, such as district, school stream (state, state-religious, Arab, Haredi and others), and town size, samples each in proportion to its share of 12th graders, and draws students at random within each stratum from a reliable list.

## Common questions

### Does a bigger sample always mean a better one?

No. A bigger sample reduces chance error, but it does not cure bias. The Literary Digest asked millions and was badly wrong in 1936, while Gallup's far smaller sample called the election correctly.

### What is the difference between random and stratified sampling?

In a simple random sample everyone has the same chance of being chosen. In a stratified sample the population is first split into groups, and each group is sampled at random in proportion to its share, which guarantees that every group appears.

### What is sampling bias?

Sampling bias is a built-in tilt in who gets into a sample, so that some kinds of people are overrepresented and others missing. It usually comes from a poor sampling frame or from letting people choose themselves.

### What is non-response bias?

Non-response bias arises when the people who do not answer a survey differ from those who do. Even a well-drawn sample becomes unrepresentative if, for example, only those with strong views reply.

### Is a convenience sample ever useful?

Yes, for trying out a questionnaire, for a first look at a new subject, or in qualitative research that seeks depth rather than numbers. Its findings simply cannot be generalized to a population.
