Populations, samples, and sampling interpretation

Learning objectives

• Distinguish target population, sampling frame and observed sample.
• Choose an appropriate sampling method for a garment style.
• Explain why representativeness beats raw sample size.

Core idea

Population = defined set of interest; frame = list from which units are drawn; sample = observed subset. Parameters describe populations; statistics estimate them.

Sampling method

• Define the target (e.g. all polos in style S123, lines A/B/C, week 12).
• Build the frame from real production records.
• Stratify by size, colour, line and shift to protect against imbalance.
• Randomize within strata; document non-response and rejects.
• Report design; do not treat a convenience sample as random.

Garment-factory example

Sampling only repaired garments from the end of a shift over-represents defects and would falsely elevate estimated defect rates. Stratified random sampling across the whole shift gives a defensible baseline.

Method

1. Define target and frame explicitly.
2. Stratify and randomize.
3. Document response, refusals and drop-outs.

Common mistakes

• Assuming a large biased sample is fine.
• Confusing frame with population.
• Using convenience samples for inference.