A non-probability sampling technique in which participants are selected because they are easily accessible to the researcher — students in the researcher's own classes, learners at a single institution, members of an existing programme. Also called opportunity or accidental sampling. It is the dominant strategy in classroom-based SLA research and language-testing pilots, where probability sampling is rarely feasible.
Convenience samples are cheap, fast, and ethically straightforward when participants are already part of the researcher's instructional context. The cost is restricted external validity: there is no statistical basis for generalising estimates to a defined population, because no member of the wider population had a known non-zero probability of inclusion. Findings hold for participants like these, in settings like this, under conditions like these — and that bound has to be stated explicitly.
Dörnyei (2007) notes that nearly all classroom-based SLA studies rely on convenience samples and argues this is acceptable provided three conditions are met: the sample is described in detail (proficiency, L1, age, instructional context), the limits on generalisation are acknowledged, and findings are interpreted as informing rather than establishing population-level claims. Replication across different convenience samples is the practical route to broader inference.
Researchers can strengthen a convenience sample by recruiting from multiple sites, ensuring within-sample variation on key learner variables (proficiency band, age, L1 background), and reporting full descriptive statistics so meta-analysts can later weight or stratify. Random assignment of an existing convenience sample to conditions — the standard quasi-experimental move — preserves internal validity for treatment effects even when external validity remains limited.
Convenience sampling is opportunistic; Purposive Sampling is intentional, selecting participants for theoretical relevance to the research question. The two often overlap in practice — a convenience sample may also satisfy purposive criteria — but the underlying logic differs.