A market researcher might use multistage cluster sampling to select a sample of consumers from all cities having a population of more than 10,000 by proceeding in stages as follows:
- First, the researcher identifies and lists all cities with populations greater than 10,000. These cities form the primary clusters.
- The researcher randomly selects a sample of these cities from the total pool.
- Within each selected city, the researcher further divides the city into smaller clusters using city blocks.
- The researcher randomly selects a sample of city blocks from within each chosen city.
- Finally, from each selected city block, individual consumers or households are randomly sampled to form the final sample.
This multistage approach effectively narrows down the sample from a large population of cities to manageable city blocks and then to individual consumers, ensuring the sample is representative while optimizing cost and logistical feasibility. City blocks act as natural clusters within cities to facilitate this cluster sampling process. This process aligns with common multistage sampling methods, where:
- Stage 1: Select cities (clusters) randomly
- Stage 2: Within chosen cities, select city blocks (sub-clusters) randomly
- Stage 3: Within selected city blocks, sample individual consumers randomly
Thus, the researcher uses successively smaller and more specific cluster units (cities, city blocks, consumers) to draw a representative sample for market research purposes.
