Pilot Deviant Case Sampling — Stress-Testing Instruments with Outlier Cases
Pilot Deviant Case Sampling · Also known as: pilot extreme case sampling, pilot outlier sampling, pilot atypical case sampling, pilot deviant-case pilot sampling
Pilot deviant case sampling is a purposive sampling strategy applied during a pilot phase in which participants are deliberately selected because they represent extreme, unusual, or atypical instances of the phenomenon under study. Rather than seeking representative participants for the pilot run, the researcher intentionally recruits outlier cases to probe the boundaries of research instruments, interview guides, or data collection protocols before the main study begins.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use pilot deviant case sampling when you are developing or validating a data collection instrument — survey, interview guide, observation checklist — and want to maximize the diagnostic value of a small pilot sample. It is particularly appropriate when the main study population spans a wide range of experience and you need to ensure the instrument works across that range. It is also useful when existing instruments are being adapted to a new context and ceiling or floor effects are suspected. Do NOT use it when the goal of the pilot is to estimate effect sizes or intraclass correlations for power analysis (which requires a representative pilot sample), or when recruiting deviant cases is not feasible due to rarity, access, or ethical constraints.
Strengths & limitations
- Maximizes diagnostic yield per pilot participant — extreme cases expose instrument flaws that average cases conceal.
- Small sample (3–10 cases) is often sufficient to identify structural weaknesses, keeping pilot costs low.
- Improves instrument validity for the full range of the target population, not just the modal respondent.
- Encourages researchers to think carefully about boundary cases and edge conditions before committing to a final instrument.
- Compatible with both qualitative and quantitative instrument development traditions.
- Deviant cases may be rare or hard to access, making recruitment slow and costly relative to a convenience pilot sample.
- Findings from the pilot cannot be used to estimate population parameters or inform power calculations for the main study.
- Researcher judgment in defining and selecting deviant cases introduces subjectivity; the definition of deviance must be transparent and documented.
- A small deviant-case pilot may miss instrument problems that are specific to the average or modal participant.
Frequently asked
How many deviant cases do I need for a pilot?
Three to ten cases is typically sufficient. The goal is coverage of extremes rather than statistical representativeness, so a small number of well-chosen outlier cases can reveal most structural instrument weaknesses. Fewer than three makes it difficult to distinguish instrument problems from individual idiosyncrasies.
Can I include pilot participants in my main study?
Generally no, especially if the instrument was substantially revised after the pilot. If the instrument was unchanged and the pilot participants meet main-study eligibility criteria, some designs allow inclusion, but this must be declared and justified in the methods section.
How do I define what counts as a deviant case?
Define deviance operationally before recruitment: specify the criterion variable and the threshold that qualifies a case as extreme (e.g., top and bottom 5% on a prior scale, record-high or record-low administrative outcomes, expert-nominated exceptional cases). Document this definition in your preregistration or methods section.
Is this appropriate for quantitative survey pilots as well as qualitative interview pilots?
Yes. In quantitative survey development, deviant-case pilots are useful for detecting ceiling and floor effects, missing response categories, and items that fail to discriminate at scale extremes. In qualitative pilots, they reveal gaps in interview guides and prompts that work for typical but not atypical participants.
How does this differ from typical case sampling used in a pilot?
Pilot typical case sampling selects participants who best represent the average or modal experience, aiming to test whether the instrument works for the most common type of respondent. Pilot deviant case sampling deliberately selects outliers to test whether the instrument works at the boundaries. The two strategies diagnose different types of instrument weakness and can be combined in a two-phase pilot.
Sources
- Patton, M. Q. (2002). Qualitative Research and Evaluation Methods (3rd ed.). Sage. [Chapter 5: Sampling Strategies, pp. 230–246 on deviant/extreme case sampling] ISBN: 978-0761919711
- van Teijlingen, E. R., & Hundley, V. (2002). The importance of pilot studies. Nursing Standard, 16(40), 33–36. link ↗
How to cite this page
ScholarGate. (2026, June 3). Pilot Deviant Case Sampling. ScholarGate. https://scholargate.app/en/survey-methodology/pilot-deviant-case-sampling
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Deviant Case SamplingSurvey Methodology↔ compare
- Purposive samplingSurvey Methodology↔ compare