What are limitations in using cluster sampling technique?

Answer 1

In a cluster sample, researchers divide subjects into strata (like cities, for example), randomly select a few strata (draw the names of a few cities from a hat) and sample every subject in those strata (question everyone in that city.)

A significant disadvantage is that you may select strata that completely overlook a feature relevant to your study.

If your study polled "What is the importance of agriculture to our country's economy?" and you questioned people from New York, Chicago, Detroit, and Los Angeles, your data may be bias because it does not include opinion from more rural areas.