When would you choose stratified sampling in a surveillance system?

• A. When you want representation across key subgroups such as age categories or geographic regions.
• B. When time constraints require only convenience samples.
• C. When subgroups are identical.
• D. When you want the fastest possible data.

Answer: (D)

Stratified sampling is used when you need representation across key subgroups and you want more precise estimates for those subgroups as well as for the overall population. By dividing the population into homogeneous groups (strata) like age categories or geographic regions and sampling from each stratum (often proportional to its size), you ensure that every important subgroup is well represented in the data. This is particularly valuable in a surveillance system where you want to compare trends or levels across different subgroups and avoid biased results that could occur if some groups were under- or over-sampled.

The idea of getting data as fast as possible isn’t what stratified sampling aims for; fastest data comes from convenience sampling or other non-probability methods. Stratification focuses on representativeness and precision across subgroups rather than speed. Similarly, if subgroups were identical, stratification wouldn’t add much value, and it wouldn’t be chosen for speed.