Skip to content
AP Statistics · Exploring One-Variable Data and Collecting Data

Cluster sampling

Cluster sampling appears in 22 reviewed AP Statistics questions and sits in Exploring One-Variable Data and Collecting Data. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Cluster sampling is tested

Example 1Harder
Researchers want to understand sleep patterns of residents of a large city (population 500,000). They randomly select 50 zip codes from the city's 200 zip codes, then randomly select 10 residents from each selected zip code. What type of sampling is this?
  1. ASimple random sampling
  2. BTwo-stage cluster samplingCorrect
  3. CCluster sampling
  4. DStratified random sampling

Why B is right

This is two-stage (multistage) cluster sampling. In the first stage, 50 zip codes (clusters) are randomly selected. In the second stage, 10 residents are randomly selected from each chosen zip code. The two-stage structure distinguishes this from simple cluster sampling, where all members of selected clusters would be included.
Why the other 3 choices miss
A
SRS would give every resident of the city an equal probability of selection. Residents in non-selected zip codes have zero probability of selection, so this is not SRS.
C
Simple cluster sampling would select zip codes and include ALL residents in those zip codes. Here, only 10 residents from each selected zip code are included — this second-stage sampling makes it multistage.
D
Stratified sampling would divide all residents into strata and sample from each stratum. All zip codes are not used here; only selected zip codes are involved.
Example 2Medium
Compared to cluster sampling, what is the main advantage of stratified random sampling in reducing sampling error?
  1. AStratified sampling always has a smaller sample size, making it cheaper.
  2. BStratified sampling represents every subgroup, giving more precise estimates.Correct
  3. CStratified sampling allows larger clusters to be selected, increasing the sample.
  4. DStratified sampling eliminates all bias, while cluster sampling introduces bias.

Why B is right

Stratified sampling ensures that all important subgroups (strata) are represented and reduces sampling variability by sampling from each stratum separately. Within each stratum, individuals are more similar to each other (lower within-stratum variability) than individuals across clusters. This typically yields more precise population estimates compared to cluster sampling, where entire clusters may be internally homogeneous but different from other clusters.
Why the other 3 choices miss
A
Incorrect because stratified sampling does not necessarily have a smaller sample size. Both methods can use the same total sample size. The advantage of stratification is precision, not sample size reduction.
C
Incorrect because stratification is about sampling from strata, not about selecting clusters. The description confuses the concepts of clusters and strata.
D
Incorrect because neither method eliminates all bias. Both can have biases (e.g., bias from poor stratum definition or cluster selection). The advantage of stratification is reduced variability, not elimination of bias.

Studied alongside

Other concepts from Exploring One-Variable Data and Collecting Data.