---
title: "Stratum (Strata) — AP Stats Definition & Exam Guide"
description: "A stratum is a subgroup of similar individuals used in stratified sampling. Learn how strata differ from clusters and how AP Stats tests them in Topic 3.3."
canonical: "https://fiveable.me/ap-stats/key-terms/stratum-strata"
type: "key-term"
subject: "AP Statistics"
unit: "Unit 1"
---

# Stratum (Strata) — AP Stats Definition & Exam Guide

## Definition

In AP Statistics, a stratum (plural: strata) is a subgroup of a population whose members share a similar characteristic, such as grade level or age group. In stratified random sampling, the population is divided into non-overlapping strata and a random sample is taken from EVERY stratum.

## What It Is

A stratum is one layer of a [population](/ap-stats/key-terms/population "fv-autolink") that you carve out before sampling. The whole point is that everyone inside a stratum is similar with respect to some [variable](/ap-stats/unit-1/language-variation-variables/study-guide/nKpeaxi1H3Ht9aFhTHKt "fv-autolink") you care about (think: freshmen, sophomores, juniors, seniors at a high school, or urban vs. rural households). The strata are non-overlapping, every individual belongs to exactly one, and together the strata cover the entire population.

Strata exist to power **stratified [random sampling](/ap-stats/unit-1/random-sampling-data-collection/study-guide/nQz8XwRMmIKKBS59qrew "fv-autolink")**, one of the named methods in Topic 3.3 (Random Sampling and Data Collection). In a stratified sample, you take a separate random sample (often an SRS) from within each stratum, so every group is guaranteed representation. Because units within a stratum are alike, the variability inside each group is small, and combining the samples gives you more precise estimates than a plain SRS of the same size. The mantra worth memorizing: strata are *similar within, different between*.

## Why It Matters

Strata live in **[Unit 3](/ap-stats/unit-3 "fv-autolink"): Collecting Data, Topic 3.3**, and they sit at the heart of two learning objectives. [AP Stats](/ap-stats "fv-autolink") 3.3.A asks you to identify a sampling method from a study description, which means spotting that 'divided the school by grade level and randomly selected 25 students from each grade' is stratified sampling, not an SRS or cluster sample. AP Stats 3.3.B asks you to explain why a method is or isn't appropriate (DAT-2.D.1 reminds you every method has trade-offs depending on the question and population). The justification for stratifying is always the same idea: when a population has groups that might respond differently, sampling from every stratum guarantees each group is represented and reduces sampling variability compared to an SRS. If you can't articulate what makes a stratum a stratum, you can't make that argument.

## Connections

### [Cluster Sample (Unit 3)](/ap-stats/key-terms/cluster-sample)

[Clusters](/ap-stats/key-terms/clusters "fv-autolink") are the mirror image of strata. A cluster should be a mini version of the whole population (different within), while a stratum is a pocket of similar individuals (similar within). In cluster sampling you randomly pick whole groups and survey everyone in them; in stratified sampling you sample from every group.

### [Simple Random Sample (Unit 3)](/ap-stats/key-terms/simple-random-sample)

[Stratified sampling](/ap-stats/key-terms/stratified-sampling "fv-autolink") is built on the SRS. Inside each stratum, you typically run a small SRS using something like a random number generator. So a stratified sample is really several SRSs stitched together, one per layer of the population.

### [Sampling Without Replacement (Unit 3)](/ap-stats/key-terms/sampling-without-replacement)

The [random selection](/ap-stats/key-terms/random-selection "fv-autolink") within each stratum is almost always done without replacement (DAT-2.C.1). Each person in a stratum can be chosen only once, which is why random-number methods tell you to ignore repeats.

### Bias and Sampling Variability (Units 3 and 5)

Stratifying doesn't fix bias, but it does shrink sampling variability. Because individuals within a stratum are alike, estimates from each layer bounce around less, so the combined estimate is more precise. That payoff connects forward to sampling distributions in Unit 5, where less variability means a tighter distribution of the statistic.

## On the AP Exam

Strata show up most often in MCQ stems that describe a study and ask you to name the sampling method (3.3.A). The giveaway phrase is 'divided into groups and randomly sampled from each group.' Watch for trap answers that flip it into cluster sampling, where whole groups are randomly chosen instead. On FRQs, the classic task is design or justification: describe how to take a stratified sample, or explain why stratifying by some variable (like grade level or region) is better than an SRS for a given question. Full credit requires the comparison logic, not just the vocabulary. Say that the groups are expected to differ on the response variable, so sampling within each stratum guarantees representation of every group and reduces variability in the estimate. Naming the method without explaining the 'similar within, different between' reasoning usually loses points.

## stratum (strata) vs Cluster

Both involve splitting a population into groups, but the logic is opposite. Strata are homogeneous (members are alike), and you randomly sample from EVERY stratum. Clusters are ideally heterogeneous (each cluster looks like a mini population), and you randomly select SOME clusters and survey everyone inside them. Quick test: if every group contributes to the sample, it's stratified; if only randomly chosen groups do, it's cluster.

## Key Takeaways

- A stratum is a non-overlapping subgroup of a population whose members share a similar characteristic, and every individual belongs to exactly one stratum.
- In stratified random sampling, you take a random sample from every single stratum, which guarantees each subgroup is represented.
- The design logic is 'similar within strata, different between strata,' which is the exact opposite of clusters, which should be different within and similar between.
- Stratifying reduces sampling variability compared to an SRS of the same size because individuals within each stratum are alike.
- On the exam, justify stratification by naming the variable you stratified on and explaining that the groups are expected to differ on the response variable.
- Stratified sampling improves precision, but it does not eliminate bias from things like bad sampling frames or nonresponse.

## FAQs

### What is a stratum in AP Stats?

A stratum is a subgroup of a population whose members share a similar characteristic, like all the juniors at a school or all households in one income bracket. In stratified random sampling, the population is split into non-overlapping strata and a random sample is drawn from each one.

### What's the difference between strata and clusters?

Strata are homogeneous groups (members are similar), and you sample from every stratum. Clusters are ideally heterogeneous mini versions of the population, and you randomly choose some clusters and survey everyone in them. If every group gets sampled, it's stratified; if only some randomly chosen groups do, it's cluster.

### Is a stratified sample better than a simple random sample?

Not automatically, it depends on the situation (that's the point of DAT-2.D.1). When the population has groups that differ on the response variable, stratifying produces less variable estimates than an SRS of the same size and guarantees every group is represented. If the groups don't actually differ, stratifying buys you little.

### Does stratified sampling remove bias?

No. Stratifying reduces sampling variability, not bias. A stratified sample can still suffer from undercoverage, nonresponse, or a bad sampling frame. Random selection within each stratum is what protects against selection bias, and even that doesn't fix nonresponse.

### How do I justify using strata on an AP Stats FRQ?

Name the stratifying variable, state that the strata are expected to differ on the response variable, and explain that randomly sampling from each stratum guarantees all groups are represented and reduces variability in the estimate. Just saying 'it's stratified' without that reasoning won't earn the explanation point.

## Related Study Guides

- [1.11 Random Sampling](/ap-stats/unit-1/random-sampling-data-collection/study-guide/nQz8XwRMmIKKBS59qrew)

## Structured Data

```json
{"@context":"https://schema.org","@graph":[{"@type":"LearningResource","@id":"https://fiveable.me/ap-stats/key-terms/stratum-strata#resource","name":"Stratum (Strata) — AP Stats Definition & Exam Guide","url":"https://fiveable.me/ap-stats/key-terms/stratum-strata","learningResourceType":"Concept explainer","educationalLevel":"AP® / High School","about":{"@id":"https://fiveable.me/ap-stats/key-terms/stratum-strata#term"},"audience":{"@type":"EducationalAudience","educationalRole":"student"},"dateModified":"2026-06-11T05:22:53.815Z","isPartOf":{"@type":"Collection","name":"AP Statistics Key Terms","url":"https://fiveable.me/ap-stats/key-terms"},"publisher":{"@type":"Organization","name":"Fiveable","url":"https://fiveable.me"}},{"@type":"DefinedTerm","@id":"https://fiveable.me/ap-stats/key-terms/stratum-strata#term","name":"stratum (strata)","description":"In AP Statistics, a stratum (plural: strata) is a subgroup of a population whose members share a similar characteristic, such as grade level or age group. In stratified random sampling, the population is divided into non-overlapping strata and a random sample is taken from EVERY stratum.","url":"https://fiveable.me/ap-stats/key-terms/stratum-strata","inDefinedTermSet":{"@type":"DefinedTermSet","name":"AP Statistics Key Terms","url":"https://fiveable.me/ap-stats/key-terms"}},{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What is a stratum in AP Stats?","acceptedAnswer":{"@type":"Answer","text":"A stratum is a subgroup of a population whose members share a similar characteristic, like all the juniors at a school or all households in one income bracket. In stratified random sampling, the population is split into non-overlapping strata and a random sample is drawn from each one."}},{"@type":"Question","name":"What's the difference between strata and clusters?","acceptedAnswer":{"@type":"Answer","text":"Strata are homogeneous groups (members are similar), and you sample from every stratum. Clusters are ideally heterogeneous mini versions of the population, and you randomly choose some clusters and survey everyone in them. If every group gets sampled, it's stratified; if only some randomly chosen groups do, it's cluster."}},{"@type":"Question","name":"Is a stratified sample better than a simple random sample?","acceptedAnswer":{"@type":"Answer","text":"Not automatically, it depends on the situation (that's the point of DAT-2.D.1). When the population has groups that differ on the response variable, stratifying produces less variable estimates than an SRS of the same size and guarantees every group is represented. If the groups don't actually differ, stratifying buys you little."}},{"@type":"Question","name":"Does stratified sampling remove bias?","acceptedAnswer":{"@type":"Answer","text":"No. Stratifying reduces sampling variability, not bias. A stratified sample can still suffer from undercoverage, nonresponse, or a bad sampling frame. Random selection within each stratum is what protects against selection bias, and even that doesn't fix nonresponse."}},{"@type":"Question","name":"How do I justify using strata on an AP Stats FRQ?","acceptedAnswer":{"@type":"Answer","text":"Name the stratifying variable, state that the strata are expected to differ on the response variable, and explain that randomly sampling from each stratum guarantees all groups are represented and reduces variability in the estimate. Just saying 'it's stratified' without that reasoning won't earn the explanation point."}}]},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"AP Statistics","item":"https://fiveable.me/ap-stats"},{"@type":"ListItem","position":2,"name":"Key Terms","item":"https://fiveable.me/ap-stats/key-terms"},{"@type":"ListItem","position":3,"name":"Unit 1","item":"https://fiveable.me/ap-stats/unit-1"},{"@type":"ListItem","position":4,"name":"stratum (strata)"}]}]}
```