Types of sampling – random, stratified, systematic and cluster MCQs With Answer

In pharmaceutics and clinical research, selecting representative samples is essential for valid results. This concise guide explains four core probability sampling methods—random sampling, stratified sampling, systematic sampling, and cluster sampling—and how they affect bias, precision, sample size, sampling frame, and representativeness in B.Pharm studies. You’ll learn when to use stratified designs to reduce variance, when systematic sampling risks periodicity, and how cluster or multistage sampling balances logistics and cost for geographically dispersed populations. Keywords: random sampling, stratified sampling, systematic sampling, cluster sampling, probability sampling, sampling frame, design effect, intra-cluster correlation, sample size calculation, representativeness. Now let’s test your knowledge with 30 MCQs on this topic.

Q1. Which statement best defines simple random sampling (SRS) in the context of pharmacoepidemiology?

• Selecting the first n individuals on a patient list
• Each member of the population has an equal and known chance of being selected
• Choosing clusters such as hospitals and sampling all patients inside
• Dividing the population into strata and sampling within each

Correct Answer: Each member of the population has an equal and known chance of being selected

Q2. What is the primary rationale for using stratified sampling in pharmacy surveys?

• To reduce data collection costs by sampling whole clusters
• To ensure representation of important subgroups and reduce overall variance
• To select every kth prescription from pharmacy records
• To replace randomization in clinical trials

Correct Answer: To ensure representation of important subgroups and reduce overall variance

Q3. How is systematic sampling typically implemented when sampling prescriptions from a large invoice list?

• Selecting clusters of consecutive prescriptions
• Choosing a random start, then selecting every kth prescription
• Stratifying invoices by drug class and sampling proportionately
• Sampling with replacement using random numbers

Correct Answer: Choosing a random start, then selecting every kth prescription

Q4. Which description best matches cluster sampling used in community pharmacy research?

• Sampling individuals uniformly across the whole population
• Randomly selecting groups (e.g., pharmacies or hospitals) and sampling within those groups
• Dividing population into strata and taking equal samples from each
• Selecting samples based on researcher convenience

Correct Answer: Randomly selecting groups (e.g., pharmacies or hospitals) and sampling within those groups

Q5. When is stratified sampling most beneficial for B.Pharm studies?

• When the population is homogeneous and spread geographically
• When cost must be minimized by sampling entire clusters
• When there are distinct subgroups with different means and within-subgroup homogeneity
• When only a convenience sample is possible

Correct Answer: When there are distinct subgroups with different means and within-subgroup homogeneity

Q6. For systematic sampling, how do you calculate the sampling interval k?

• k = required sample size / population size
• k = population size / required sample size
• k = square root of population size
• k = number of strata

Correct Answer: k = population size / required sample size

Q7. What is a key risk when applying systematic sampling to pharmacy dispensing logs?

• Inability to compute sample weights
• Introduction of periodicity bias if a pattern aligns with the sampling interval
• Excessive within-cluster variance always lowering precision
• Requirement for a complete sampling frame of clusters only

Correct Answer: Introduction of periodicity bias if a pattern aligns with the sampling interval

Q8. In cluster sampling with unequal cluster sizes, which technique helps reduce selection bias?

• Stratifying clusters by outcome
• Sampling with replacement
• Probability-proportional-to-size (PPS) sampling
• Systematic sampling within each cluster

Correct Answer: Probability-proportional-to-size (PPS) sampling

Q9. The design effect (DEFF) in cluster sampling is primarily influenced by which parameter?

• Survey response rate only
• Intracluster correlation coefficient (ICC) and cluster size
• Number of strata only
• Whether sampling is with or without replacement

Correct Answer: Intracluster correlation coefficient (ICC) and cluster size

Q10. Which scenario most justifies using cluster or multistage sampling in a nationwide pharmacy practice study?

• When a complete list of every patient is available
• When the outcome variance is negligible across subgroups
• When the population is geographically dispersed and a full sampling frame of individuals is impractical
• When precise estimates for small subgroups are required without weighting

Correct Answer: When the population is geographically dispersed and a full sampling frame of individuals is impractical

Q11. What does proportionate stratified sampling mean in relation to stratum sizes?

• Each stratum contributes the same sample size regardless of population size
• Sample sizes are allocated in proportion to each stratum’s population size
• Only the largest stratum is sampled
• Samples are chosen to maximize between-stratum homogeneity

Correct Answer: Sample sizes are allocated in proportion to each stratum’s population size

Q12. Which allocation strategy gives more samples to strata with higher variability to improve precision?

• Proportionate allocation
• Disproportionate random allocation
• Optimal (Neyman) allocation
• Systematic allocation

Correct Answer: Optimal (Neyman) allocation

Q13. In simple random sampling without replacement (SRSWOR), how does selecting one individual affect the remaining selection probabilities?

• Probabilities remain unchanged because replacement occurs
• Probabilities increase for the remaining units because the population reduces by one
• Probabilities decrease uniformly for all remaining units
• Selection becomes stratified automatically

Correct Answer: Probabilities increase for the remaining units because the population reduces by one

Q14. For a national pharmacoepidemiology survey where listing all individuals is impossible, which sampling design is most commonly used?

• Simple random sampling of individuals from an exhaustive frame
• Multistage cluster sampling with clusters selected at higher administrative levels
• Single-stage systematic sampling with small k
• Convenience sampling of accessible clinics only

Correct Answer: Multistage cluster sampling with clusters selected at higher administrative levels

Q15. What is a sampling frame in the context of clinical pharmacy research?

• The statistical software used to analyze samples
• A list or database of all elements from which the sample is drawn
• The final report describing sampling results
• An arbitrary grouping of clusters for convenience

Correct Answer: A list or database of all elements from which the sample is drawn

Q16. If a sampling frame excludes private pharmacies when the target is all pharmacies, which bias is introduced?

• Measurement bias
• Coverage bias
• Confounding bias
• Observer bias

Correct Answer: Coverage bias

Q17. For estimating a population mean with limited resources, which sampling method often yields lower variance than SRS for the same overall sample size?

• Cluster sampling with large clusters
• Stratified sampling with homogeneous strata
• Convenience sampling
• Systematic sampling without random start

Correct Answer: Stratified sampling with homogeneous strata

Q18. What does multistage sampling combine in large-scale pharmacoepidemiological studies?

• Only systematic sampling at each stage
• Cluster selection at primary stages and random or systematic sampling within selected clusters
• Complete enumeration of all units in selected clusters only
• Non-probability methods for speed

Correct Answer: Cluster selection at primary stages and random or systematic sampling within selected clusters

Q19. Which tool is recommended for implementing true simple random sampling when a sampling frame exists?

• Random number generator or random number table
• Choosing the first entries in an alphabetized list
• Sampling every record from a single clinic only
• Allocating samples by researcher preference

Correct Answer: Random number generator or random number table

Q20. When comparing medication adherence between males and females, which sampling approach ensures sufficient representation of both sexes?

• Cluster sampling only
• Stratified sampling by sex with appropriate allocation
• Systematic sampling without considering sex
• Convenience sampling at a single hospital

Correct Answer: Stratified sampling by sex with appropriate allocation

Q21. In a cluster sample where primary sampling units are hospitals, what is typically sampled at the next stage?

• The entire national registry again
• Patients or prescriptions within the selected hospitals
• Only hospital administrators for interviews
• Random clusters of drug classes

Correct Answer: Patients or prescriptions within the selected hospitals

Q22. For systematic sampling, what is essential before selecting every kth element?

• Sorting the frame by outcome variable
• Choosing a random start between 1 and k
• Ensuring clusters are homogeneous
• Applying weight adjustments

Correct Answer: Choosing a random start between 1 and k

Q23. How does a high intracluster correlation (ICC) affect the efficiency of cluster sampling?

• Increases efficiency because clusters are diverse
• Decreases efficiency because observations within clusters are similar
• Has no effect on sample variance
• Automatically reduces required sample size

Correct Answer: Decreases efficiency because observations within clusters are similar

Q24. Which of the following is a non-probability sampling method often contrasted with the four probability methods discussed?

• Simple random sampling
• Purposive (judgmental) sampling
• Stratified random sampling
• Systematic sampling with random start

Correct Answer: Purposive (judgmental) sampling

Q25. When planning a cluster-randomized field trial in community pharmacies, what must be included in sample size calculations?

• Only the number of individual participants without adjustments
• An adjustment for design effect due to clustering
• No need for power calculations in cluster trials
• Only the ICC but not cluster size

Correct Answer: An adjustment for design effect due to clustering

Q26. If disproportionate stratified sampling is used, what analytical step is often required during estimation?

• Ignore stratum sizes to simplify analysis
• Apply sampling weights to obtain unbiased population estimates
• Convert the sample into a cluster sample
• Use periodic sampling to adjust for bias

Correct Answer: Apply sampling weights to obtain unbiased population estimates

Q27. Selecting every 10th prescription from a pharmacy log after a random start is an example of which method?

• Stratified sampling
• Systematic sampling
• Cluster sampling
• Purposive sampling

Correct Answer: Systematic sampling

Q28. Which sampling approach is most appropriate when the target population is small, well-defined, and a complete list exists?

• Simple random sampling without replacement
• Multistage cluster sampling
• Convenience sampling
• Purposive sampling

Correct Answer: Simple random sampling without replacement

Q29. If strata boundaries or sizes are unknown and difficult to define, which approach is often more practical?

• Proportionate stratified sampling
• Cluster or multistage sampling with appropriate adjustments
• Exact Neyman allocation
• Ensuring equal samples per stratum regardless of size

Correct Answer: Cluster or multistage sampling with appropriate adjustments

Q30. For a large, geographically dispersed population where minimizing field costs is the priority, which sampling method is usually preferred?

• Simple random sampling of individuals across the country
• Cluster sampling or multistage sampling to reduce travel and listing costs
• Disproportionate stratified sampling without weighting
• Systematic sampling with k = 1

Correct Answer: Cluster sampling or multistage sampling to reduce travel and listing costs

Download