Anchor Case Library¶
Artifact — instantiates Dense-Subset Coverage Design
Maintains representative-by-proximity cases, exemplars, prototypes, personas, benchmarks, or scenarios with declared coverage scope.
An anchor case library is the curated, human-readable repository of representative cases — exemplars, prototypes, personas, benchmarks, scenarios — where each entry carries an explicit declared coverage scope: a statement of which real situations it stands in for and how far that claim reaches. Its defining move is to make the representation claim legible and bounded, entry by entry. Where an epsilon-net expresses coverage as a formal metric radius, the case library expresses it in the domain's own language: this persona speaks for these users, this benchmark stands for this class of production inputs, this prototype represents this corner of the use space — and no further. It is deliberately the informal cousin of the metric constructions: proximity is judged qualitatively, by argued similarity rather than a computed distance, which is what lets it operate in spaces too rich or too human to metrize, at the cost of the guarantee.
Example¶
A bank's product team maintains a persona library for its mobile app. Rather than claiming to model every customer, the library holds eight anchor personas, each with a declared scope: "Maya, gig-economy earner with irregular income" explicitly stands for customers with volatile cash flow and thin credit history; "Walter, retiree managing fixed income and accessibility needs" stands for another region of the user space. Each persona names the situations it represents and, just as importantly, the situations it does not — Maya's scope note says nothing about small-business owners, so a designer cannot quietly stretch her to cover them. When a new feature idea arrives — say, a shared household budget — the team asks which persona is nearest. If a candidate feature maps cleanly onto Walter's or Maya's declared scope, they design against that anchor. If a proposed use — a teenager's first supervised account — sits far from every persona's scope, that distance is itself the signal: the library has no anchor here, and the team must either add one or admit the design is flying blind. The scope declarations are what keep the eight personas from silently over-claiming to cover a customer base of millions.
How it works¶
- Curate exemplars, not samples. Choose a small set of vivid, concrete cases picked to represent regions of the space, each fleshed out enough to design, test, or reason against.
- Declare each scope. Attach to every entry an explicit statement of what it stands for and the edge of that claim — the entry's own local boundary within the larger target space.
- Set the "close enough" bar. Say how similar a real situation must be to an exemplar to count as represented by it, in domain terms rather than a coordinate distance.
- Provide the lookup. Give users a way to find the nearest exemplar for a situation at hand — and to notice when nothing is near enough.
Tuning parameters¶
- Library size — how many exemplars to maintain. More cases cover the space more finely but dilute each one's memorability and raise upkeep; too few and each is stretched past its honest scope.
- Scope breadth per entry — how wide a region each exemplar claims. Broad scopes need fewer cases but invite over-reach; narrow scopes are honest but leave more of the space unclaimed.
- Richness vs. count — deep, fully-realized exemplars (memorable, expensive) vs. many thin ones (cheap, forgettable).
- Representation basis — whether entries represent by typicality (the common case) or by spread (deliberately including extremes and edges), which changes what the middle of the library means.
- Scope-language rigor — how formally the "stands for / does not stand for" notes are written, trading auditability against authoring cost.
When it helps, and when it misleads¶
Its strength is that it puts a bounded, communicable face on a coverage claim: a persona or benchmark with a declared scope tells a whole team both what it covers and where its authority stops, which is exactly what turns a pile of examples into a cover with visible edges. It thrives in spaces — user experience, policy, design — where a behavior-preserving metric is unavailable but argued similarity is still meaningful.
It misleads when the exemplars are quietly promoted from representative-by-proximity to statistically representative. A persona[n1] library is a coverage device, not a census: eight personas may span the important regions of a user space while matching none of its demographic proportions, and treating "we have a persona for that" as "our users are mostly like this" is the classic misuse, the archetype's representativeness overclaim. Scope notes also rot — a persona written three years ago may no longer describe the segment it names — and qualitative proximity invites wishful stretching, designers reaching an exemplar past its declared edge because it is the nearest one they have. The guarding discipline is to keep each scope declaration explicit and current, to refuse to stretch an exemplar past its stated edge, and to treat a situation far from every entry as uncovered rather than as "close enough to Maya."
How it implements the components¶
dense_reference_subset— the library is the anchor set, rendered as concrete, human-legible exemplars rather than abstract points.target_space_boundary— each entry's declared coverage scope bounds what it stands for, and the union of scopes makes the library's overall reach explicit.coverage_tolerance— the "close enough to count as represented" bar, stated in domain terms, sets how far an exemplar's claim extends.anchor_to_target_mapping_rule— the lookup that pairs a situation at hand with its nearest exemplar, or reports that none is near enough.
It does not compute a formal distance metric or a covering-radius guarantee (that is Epsilon-Net or Covering Grid), render a gap assessment across axes (that is Coverage Heatmap), or keep the library current as the domain drifts (that is Adaptive Refinement Loop).
Related¶
- Instantiates: Dense-Subset Coverage Design — it is the human-curated, scope-declaring anchor set for spaces that resist a formal metric.
- Consumes: Adaptive Refinement Loop keeps the library's entries and scopes from going stale as the domain changes.
- Sibling mechanisms: Epsilon-Net or Covering Grid · Space-Filling Design · Boundary-Value Test Suite · Coverage Heatmap · Nearest-Neighbor Assignment Rule · Adaptive Refinement Loop · Sensor or Service Radius Map
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Maintains representative-by-proximity cases, exemplars, prototypes, personas, benchmarks, or scenarios with declared coverage scope, making its operative form a non-executable information artifact that externalizes static or prospective structure.
Independent corroboration: The frozen evidence defines Anchor Case Library as 'Maintains representative-by-proximity cases, exemplars, prototypes, personas, benchmarks, or scenarios with declared coverage scope', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Human-Computer Interaction
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Persona and scenario libraries in user-centered design most directly embody a curated set of concrete cases used to cover a heterogeneous use space.
Related originating lineages:
- Data Science & Analytics — Benchmark datasets and canonical evaluation cases provide a material library-based comparison lineage.
- Education & Pedagogy — Worked examples and exemplar libraries support transfer.
- Engineering & Design — Reference designs, prototypes, and test cases provide parallel anchor artifacts.
- Law & Governance — Case libraries and precedent provide professionally curated anchors.
- Library & Information Science — Curated case collections and explicit scope metadata supply the repository discipline.
- Statistics & Experimental Design — Design-point and benchmark selection contributes the coverage logic.
Review resolution: HCI research uses evidence-based personas, scenarios, and user stories as durable representations that preserve user and workflow characteristics for design. That is the closest lineage for a scoped anchor-case library, while benchmarks, exemplars, prototypes, precedents, and knowledge organization materially broaden the synthetic artifact.
Attribution caveat: The generalized library deliberately unifies personas, scenarios, exemplars, benchmarks, prototypes, and precedents that have distinct professional histories.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
- Turner, Reeder, and Ramey — Scenarios, Personas and User Stories
- Persona Cases: A Technique for Grounding Personas
Notes¶
The library shares the nearest-exemplar lookup with the Nearest-Neighbor Assignment Rule, but the two live at different layers: the assignment rule is the algorithm that gates a match on a computed distance, while the library is the artifact whose entries carry declared, argued scopes. A case library can be routed by a nearest-neighbor rule, but its coverage claim rests on human-written scope notes, not on a metric.
[n1] A persona is a fictional but grounded archetype of a user, built to stand in for a segment of real users during design. Personas are a coverage device — chosen to span the important regions of a user space — and are explicitly not a statistical sample; treating a small persona set as demographically representative is a well-known misuse. ↩