Stratified sampling is to divide all elements into several types or levels according to certain characteristics or signs (such as age, gender, occupation or region, etc.). ), and using the first two methods to extract a sub-sample, all sub-samples constitute the total sample.
discrepancy
Simple random sampling operation is simple, but the total number is too large and the workload is too large.
Stratified sampling fully ensures the consistency of sample structure and population, and improves the representativeness of samples. It is not applicable when the overall difference is not obvious, and it needs to be used in combination with other sampling methods.