sampling-strategies

# Sampling Strategies in Qualitative Research Qualitative sampling prioritizes **information-rich cases** and **analytic progression** over statistical representativeness. Your sampling chapter should explain logic, not pretend to randomize for generalization in a positivist sense. ## Purposeful sampling (Patton) Purposeful sampling selects cases that illuminate the phenomenon intensely. Common variants: - **Maximum variation:** Deliberately include diverse contexts to surface core patterns and boundaries. - **Homogeneous:** Deepen understanding within a similar subset when noise reduction helps early theory. - **Typical case:** Clarify the “usual” pattern when stakeholders misunderstand extremes. - **Critical case:** A instance that can falsify or strongly test an emerging claim. - **Criterion:** All participants meet explicit inclusion criteria tied to the research question. - **Snowball/chain:** Referrals lead to new participants; efficient for hidden populations. ## Theoretical sampling (Glaser) After initial coding, **theoretical sampling** seeks data to elaborate **properties, dimensions, conditions, and consequences** of emerging categories. You sample for **analytic gaps**, not demographic quotas (unless demographics are theoretically relevant). The question is: “What data do I need next to develop this category?” ## Snowball and chain referral sampling Use when populations are hard to reach or trust is network-based. Mitigate bias from seed participants by using **multiple entry points** and monitoring whose voices dominate. Document referral chains in audit materials. ## Maximum variation vs typical case Maximum variation guards against **provincial theory**; typical case helps communicate **mundane patterning**. In GT, these are tools in service of emergent categories—choose based on what the evolving theory demands. ## Convenience sampling (with caveats) Convenience sampling is acceptable only when limitations are **transparent** and you take compensatory steps (prolonged engagement, triangulation, negative case search). Never dress convenience up as theoretical sampling without analytic justification. ## Sample size in qualitative research “How many?” depends on **study scope, data richness, and saturation**. A focused GT project may achieve workable saturation with fewer cases if incidents are dense; a comparative multi-site study may need many more. Report **saturation reasoning** and **disconfirming efforts**, not a fake power analysis. ## Saturation-based sampling Saturation means **no new properties/dimensions** are emerging for a category (in the Glaserian sense), not mere repetition of wordings. Track saturation **by category**, not only globally—some categories saturate early while others remain thin. ## Integration tip In proposals, distinguish **initial purposeful sampling** (who can get you started) from **later theoretical sampling** (who/what you need next as categories form). This shows methodological maturity in GT. ## Checklist - [ ] Sampling logic matches methodology (purposeful vs theoretical). - [ ] Inclusion/exclusion criteria explicit and ethically justified. - [ ] Referral/snowball bias considered and mitigated where possible. - [ ] Saturation tracked with documented negative case searches. - [ ] Limitations of non-probability sampling frankly acknowledged. ## References (starting points) - Patton, M. Q. *Qualitative research and evaluation methods* — purposeful sampling typology. - Glaser, B. G., & Strauss, A. L. *The Discovery of Grounded Theory* — theoretical sampling origins. - Emmel, N. *Sampling and choosing cases in qualitative research*.