Follow-Up Case Sampling Plan¶
Sampling plan — instantiates Deviant Case Analysis
Selects follow-up cases, audits, or monitoring triggers to test whether the revision prompted by a deviant case is isolated, subgroup-specific, or broadly general.
A revision drawn from one deviant case is a hypothesis about how far the exception reaches — and that reach is exactly what a single case cannot settle. The Follow-Up Case Sampling Plan is the forward-looking step that decides where to look next: given a candidate revision, it scans the remaining unexplained cases and designs a deliberate sample of further cases, tests, audits, or monitoring triggers whose results will locate the revision on the spectrum from isolated one-off to subgroup rule to general law. Its defining orientation is outward and downstream: it consumes a proposed revision and produces a plan for new evidence, driven by the leftover residual population rather than by the deliberation that produced the revision.
Example¶
A conservation model predicts steady decline for a river basin's native trout under warming summers, yet at one spring-fed tributary the population unexpectedly rebounds. A revision is proposed: cold groundwater upwelling buffers the temperature the model assumes. The sampling plan turns that proposal into a test. It first scans the basin's residuals — which other tributaries are also over-performing the decline model — and then designs a stratified follow-up sample: six groundwater-fed tributaries to test the subgroup hypothesis, six surface-fed tributaries matched on elevation and flow to serve as the disconfirming contrast, and a monitoring trigger that re-samples any site whose summer temperature crosses a set threshold. The plan states in advance what result would widen the revision to all cold-buffered streams and what result would retract it to this one anomalous creek. The output is not a conclusion but a designed pathway to one, so the revision is neither over-generalized from a single creek nor dismissed as a fluke.
How it works¶
- State the generalization question. Is the revision being tested as isolated, subgroup-specific, or general? That choice sets how wide to sample.
- Scan the residuals for candidates. Sweep the corpus for other cases consistent — or pointedly inconsistent — with the revision, and let that scan populate the sampling frame.
- Design a sample that can break the revision. Include confirming subgroup cases and deliberately disconfirming or contrast cases, plus monitoring triggers for cases not yet observed.
- Pre-commit the decision rule. Write down which results widen the revision and which retract it, before the follow-up data arrive.
Tuning parameters¶
- Generalization target — isolated, subgroup, or population. The wider the claim, the wider and more varied the sample must be.
- Confirm/disconfirm balance — how many follow-ups are chosen to break the revision versus confirm it. A disconfirm-heavy plan is more honest and harder to arrange.
- Residual-scan reach — how far across the corpus the scan looks for candidate follow-ups. Wider reach catches distant analogues but dilutes effort.
- Trigger sensitivity — the thresholds at which monitoring re-samples a case; sensitive triggers catch drift early but fire often.
When it helps, and when it misleads¶
Its strength is that it stops a one-case revision from being either over-generalized or waved away, by putting it to a designed test on genuinely new evidence. Its honest failure mode is confirmation sampling — quietly choosing follow-ups likely to agree, so the revision "generalizes" by construction rather than by test; a plan that only ever samples where it expects to win learns nothing. The classic misuse is sampling until a threshold is crossed and then stopping, which manufactures support out of optional stopping. The guarding discipline is to pre-commit both the disconfirming cases and the retract condition, in the spirit of theoretical sampling — choosing the next cases to develop and stress the emerging idea, not to decorate it.[n1]
How it implements the components¶
The plan realizes the archetype's test-the-reach step — the machinery that converts a case-level revision into a generalization question with a sampling answer:
follow_up_case_sampling_plan— it is the plan itself: which further cases, tests, audits, and monitoring triggers to pursue, chosen to locate the revision on the isolated-to-general spectrum.residual_pattern_scan— it scans the remaining unexplained cases to decide where the next evidence should come from, using the leftover residuals as the sampling frame.
It does not arrange the matched pairs for the current comparison (matched_or_contrast_comparison_set — the Matched Case Pairing Matrix), record the revision it is testing (theory_revision_register — the Boundary Condition Revision Workshop), or map where the pattern holds (boundary_condition_map — likewise the workshop).
Related¶
- Instantiates: Deviant Case Analysis — supplies the archetype's generalization-testing step.
- Consumes: Boundary Condition Revision Workshop for the revision under test, and the Deviation Residual Table for the residual population it samples from. Where that table displays one case's gap, this plan scans the residual population to choose the next cases.
- Sibling mechanisms: Deviant Case Selection Protocol · Deviation Residual Table · Matched Case Pairing Matrix · Omitted Variable Probe · Within-Case Process-Tracing Memo · Boundary Condition Revision Workshop · Rival Explanation Elimination Table
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Follow-Up Case Sampling Plan operates as a non-executable information artifact that externalizes static or prospective structure because it selects follow-up cases, audits, or monitoring triggers to test whether the revision prompted by a deviant case is isolated, subgroup-specific, or broadly general.
Independent corroboration: The frozen evidence defines Follow-Up Case Sampling Plan as 'Selects follow-up cases, audits, or monitoring triggers to test whether the revision prompted by a deviant case is isolated, subgroup-specific, or broadly general', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Ethnography & Qualitative Methods
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Purposeful selection of deviant and follow-up cases is rooted in qualitative case-study and grounded-theory methods.
Related originating lineages:
- Biology & Ecology — Ecological contrast-site sampling materially shapes the page's subgroup and residual scan.
- Statistics & Experimental Design — Sampling design and subgroup validation materially shape judgments about generality.
Review resolution: Both reviewers agree that ethnography_qualitative_methods is primary. I retain statistics_experimental_design, biology_ecology only as formative origin lineage(s), without treating every later application as an origin. cross_disciplinary_synthesis is appropriate because the exact artifact combines contributions from multiple professional lineages. Reach is multi_domain as a separate applicability judgment: it does not widen or narrow the recorded provenance. Encyclopedia synthesis is true because the exact generalized artifact is an encyclopedia-authored combination or refinement. The secondary differences are reconciled with no unresolved primary-provenance ambiguity.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Theoretical sampling — the grounded-theory practice (Glaser & Strauss) of selecting each next case specifically to develop, saturate, or challenge an emerging concept, rather than to achieve statistical representativeness. Follow-up sampling after a deviant case is this idea applied to test how far a revision travels. ↩