BB2025 Chap.2 Sampling and Experimental Study Design
Sampling and Experimental Study Design
Classify studies from investigator control
An experiment assigns an intervention or condition. A quasi-experiment introduces a change without random individual allocation. An observational study records exposures that already exist.
Timing and comparison then refine the design: cohorts move from exposure toward outcomes, case-control studies begin with outcome status and reconstruct exposure, and cross-sectional studies observe variables within one period.
These labels describe how evidence is formed rather than ranking quality automatically.
Find the counterfactual
The comparison group or comparison period represents what would have happened without the exposure. Random allocation strengthens that representation by balancing baseline causes in expectation.
Controlled time-series and natural experiments can provide useful alternatives when randomisation is infeasible, but they require a clear intervention point, stable measurement and evidence that concurrent changes do not explain the result.
Sampling controls the reach of the claim
The target population, accessible population and recruited sample are different groups.
Probability sampling gives known selection chances when a usable frame exists. Stratified, cluster and multistage methods trade field efficiency against variance and design complexity. Non-response and volunteer entry can still change the analysed sample, so response processes and population boundaries belong in the interpretation.
What this chapter covers
- 01
Experimental and observational assignment
- 02
Randomised and quasi-experimental designs
- 03
Cohort, case-control and cross-sectional timing
- 04
Counterfactual comparisons
- 05
Allocation and analysis units
- 06
Probability sampling frames
- 07
Stratified and cluster sampling
- 08
Non-response and population coverage
Worked example · free
Classify a controlled service change
- 2Identify whether exposure was assigned randomly.
- 2Name the design from assignment and time.
- 4Explain the contemporaneous comparison and remaining threat.
Key terms
- Random Allocation
- Assignment by chance so baseline causes are balanced between intervention groups in expectation.
- Quasi-experiment
- An intervention study that constructs a comparison without random individual allocation.
- Cohort Study
- An observational design that begins with exposure groups and follows them toward outcomes.
- Case-control Study
- An observational design that selects by outcome status and reconstructs earlier exposure.
- Sampling Frame
- The operational list or structure from which eligible population units can be selected.
- Cluster Sampling
- Selection of natural groups before individuals, reducing field cost while increasing within-group dependence.
Sampling and Experimental Study Design FAQ
What is the first question when identifying a study design?
Ask who determined exposure or intervention. Investigator assignment points toward an experiment, while naturally occurring or chosen exposure points toward observation. Then inspect whether allocation was random, how the comparison was formed and whether observations move forward, backward or occupy one period. This reconstruction is safer than matching a familiar label.
Does random allocation make a sample representative?
No. Allocation supports comparison between study groups after recruitment. Sampling supports inference from the recruited sample to a target population. A trial can balance its groups well while enrolling a narrow volunteer population, and a representative survey can describe prevalence accurately without supporting a causal treatment conclusion.
Why does the analysis unit matter?
The unit receiving allocation or exposure may be a clinic, class or community even when outcomes are recorded for individuals. People within the same unit share procedures and context, so treating every record as independent exaggerates information. Record the allocation, observation and analysis units when planning sample size and modelling dependence.
Exam move
Practise design identification from plain descriptions with the formal label hidden. Write four answers for each: who assigned exposure, what represents the comparison, which direction time runs and what the independent unit is. Draw one timeline for cohort, case-control, cross-sectional and interrupted designs.
For sampling, map the target population to the frame, invitation, response and analysed sample, then identify where coverage or self-selection changes the population to which the result can travel. Add one realistic threat beside every design and one control that must operate before analysis. For cluster and repeated designs, write the allocation, observation and analysis units on separate lines.
End each exercise with a transport sentence naming the target group and the feature most likely to modify the effect outside the study setting. Reclassify the same scenario after changing one verb from assigned to chose; noticing how that single word changes design strength is excellent final-test practice. For every randomised example, identify concealment before assignment and masking after assignment as separate safeguards.
Then explain which bias each safeguard addresses. Treating them as interchangeable hides exactly where knowledge can alter entry or measurement.
Working through Sampling and Experimental Study Design in BB2025? Sia is AskSia’s AI Statistics tutor — ask any BB2025 Sampling and Experimental Study Design question and get a clear, step-by-step explanation grounded in how BB2025 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
