Which one of the following is a non-probability sampling method?

Concept

Sampling designs fall into two families, and the dividing line is a single criterion: whether the units that enter the sample are selected by a chance mechanism that gives every unit of the population a known, non-zero probability of being selected. A design that meets this criterion belongs to the probability (random) family; a design in which the investigator's judgement, convenience, or a category count decides who enters belongs to the non-probability family.

The practical consequence follows straight from the definition: only when selection probabilities are known can the sampling error be estimated and the finding be generalised to the population with a stated confidence level.

Applying the criterion

1. Look for a randomisation device — a lottery draw, a random-number table, or an equivalent chance procedure — and ask at which stage of the design it operates.

2. Simple random sampling operates that device on individual units taken from a complete list of the population, so each unit's selection probability is fixed by chance and is known.

3. Stratified sampling first divides the population into homogeneous strata and then operates the same device inside each stratum; the chance procedure is retained and only its scope is narrowed.

4. Cluster sampling divides the population into naturally occurring groups and operates the device on the groups; the chance procedure is retained and only the unit of selection changes from a person to a cluster.

5. Quota sampling fixes the number of respondents required in each category in advance and leaves the field investigator free to fill those counts with whoever is accessible and matches the category, so no chance procedure operates at any stage and the probability of a given person entering the sample is unknown.

Cross-check

Test the definition from the opposite direction. Quota sampling is often described as the non-probability counterpart of stratified sampling, because both fix the composition of the sample by category in advance. The single difference is what happens inside a category: stratified sampling still selects at random there, while quota sampling selects by accessibility. That one substitution is exactly what moves a design out of the probability family, which confirms the classification.

• Probability family: simple random, stratified, cluster, systematic and multi-stage sampling.

• Non-probability family: quota, convenience (accidental), purposive (judgement) and snowball sampling.

Quota sampling is therefore the design in this set that sits outside the probability family, which makes it the non-probability sampling method.