Tiered Decision Support¶
Software tool — instantiates Progressive Disclosure
A decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions.
Tiered Decision Support presents a recommendation or conclusion first, then lets the user descend into the reasoning behind it — the rationale, the evidence, the assumptions, and the exceptions that would overturn it — so a decision-maker can act on the summary yet audit it when the stakes warrant. Its defining idea is that the layers are tiers of justification for a single judgment: recommendation, then why, then the evidence, with the caveat that would reverse the recommendation deliberately pulled to the front. Unlike a mechanism that hides its supporting detail for tidiness, this one keeps the reversing exception visible precisely so the confident summary is never mistaken for the whole truth.
Example¶
A clinical decision-support tool, given a patient's chart, leads with a recommendation: "Start drug X, 50 mg daily," alongside a confidence indicator. A busy clinician treating a routine case can act on that line. One who wants to check drills into the next tier — the rationale, tying the recommendation to the relevant guideline and this patient's labs — and, below it, the evidence tier citing the trial criteria the guideline rests on. Above all of it, a red exception banner is fixed in the top tier, not buried: "Contraindicated if eGFR < 30; this patient is 28 — consider drug Y." The clinician acts fast on the headline for the ordinary case, inspects the basis when something feels off, and cannot miss the contraindication that would flip the decision, because the tool refuses to bury it.
How it works¶
The tool leads with the recommendation as the default layer, and stacks tiers of justification beneath it — rationale, then evidence, then the assumptions the recommendation depends on — each reachable by drilling in. The move that distinguishes it from an ordinary summary-then-detail tool is where it puts the exceptions: any caveat that would reverse the recommendation is promoted to the top tier and shown up front, rather than left for a user who drills deep enough to find it. Uncertainty is shown, not hidden, so the summary reads as a recommendation to interrogate rather than a verdict to obey. The layers are not records or data levels; they are the reasons for one judgment, ordered from conclusion down to its foundations.
Tuning parameters¶
- Confidence display — how prominently the recommendation's certainty is shown, from a bare label to a full uncertainty range. More candor invites scrutiny; too much can paralyze routine action.
- Drill depth — how far the evidence tier goes, from a one-line "per guideline X" to the underlying studies and data. Deeper serves auditors; it also raises the cost of building and maintaining the tiers.
- Exception prominence — how forcefully a reversing caveat is surfaced in the top tier. Louder guards against missed contraindications; overused, it desensitizes.
- Assumption transparency — whether the recommendation's premises are exposed as an inspectable tier or left implicit. Exposed premises are auditable but add surface area.
- Override friction — how much the tool asks when a user departs from the recommendation, balancing accountability against autonomy.
When it helps, and when it misleads¶
Its strength is combining fast action with auditability: the routine case is handled from the headline, while the reasoning stays one drill away for the case that deserves scrutiny — and the reversing exception is kept in view so a summary is never trusted blindly.
Its failure mode is false confidence. A clean, authoritative recommendation invites automation bias — users stop checking and defer to the tool even when its basis is thin or its inputs are wrong — and the tidier the summary tier, the stronger that pull.[n1] A second failure is an evidence tier that looks inspectable but is not genuinely falsifiable, so drilling in reassures without informing. The classic misuse is demoting the contraindication to a lower tier for a cleaner headline — exactly the detail that must stay up front. The guarding discipline is to make the reversing exception a first-tier element, keep the evidence tier genuinely inspectable, and show uncertainty rather than smoothing it away, so the summary earns trust instead of merely inviting it.
How it implements the components¶
default_summary— the recommendation (with its confidence) is the first layer, the thing a user can act on without drilling.critical_detail_exception— any caveat that would reverse the recommendation — a contraindication, a disqualifying condition — is promoted to the top tier and shown before the user acts, not left to be discovered by drilling.drilldown_path— the user descends from the recommendation into tiers of rationale, evidence, and assumptions to interrogate the basis of the judgment.
It does NOT implement return_path — the browse-a-list-and-come-back navigation of Summary-Detail View; tiered decision support is not a set of records to browse but a single recommendation to interrogate. Nor does it use orientation_cue or depth_label for a data hierarchy the way Drill-Down Dashboard does — its tiers are the *reasons for one judgment, not aggregation levels.*
Related¶
- Instantiates: Progressive Disclosure — it is the archetype's decision face: a recommendation up front, its justification and reversing exceptions preserved for inspection.
- Sibling mechanisms: Summary-Detail View · Drill-Down Dashboard · Expandable Section · Just-in-Time Help · Advanced Settings Panel · Staged Onboarding · Progressive Training Module · Wizard or Stepper Workflow
Editorial Notes¶
Form Classification¶
Form family: Interface, Display & Cue
Rationale: Tiered Decision Support operates as a user-facing prompt, display, template, or perceptual cue that shapes attention and action at the point of use because it a decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions.
Independent corroboration: The frozen evidence defines Tiered Decision Support as 'A decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions', so its operative form is Interface, Display & Cue.
Nearest alternative: Decision, Gate & Allocation — Tiered Decision Support includes features of a case-specific gate, selection, routing, prioritization, or resource disposition, but its defining operation is a user-facing prompt, display, template, or perceptual cue that shapes attention and action at the point of use.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Human-Computer Interaction
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Tiered decision support derives most directly from human-computer interaction's usability, wayfinding, and information-design tradition; its defining operation is to a decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions.
Related originating lineages:
- Architecture & Urban Planning — Planning's spatial allocation, accessibility, and territorial-design tradition provides a formative adjacent lineage for the same tiered decision support operation.
- Computer Science & Software Engineering — Computer science and software-engineering practice supplies a parallel or contributing lineage for the mechanism's defining operation: a decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions.
- Psychology — Experimental, clinical, and behavioral psychology supplies a parallel or contributing lineage for the mechanism's defining operation: a decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions.
Review resolution: Both blind reviewers independently select human_computer_interaction as the primary historical origin for the concrete operation—A decision-support mechanism that presents a recommendation first and allows deeper inspection of rationale, evidence, assumptions, and exceptions. The queued differences concern alternate origin disagreement, origin mode disagreement, encyclopedia synthesis disagreement, not the primary lineage. I retain every alternate that either reviewer explains, without a numeric cap, and choose origin_mode=cross_disciplinary_synthesis because the reviewers' combined evidence identifies material construction from multiple disciplines. domain_reach=universal records later portability rather than multiplying historical origins; confidence=high is the conservative shared evidentiary level, and encyclopedia_synthesis=true preserves either reviewer's affirmative synthesis finding.
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] Automation bias: the tendency to over-rely on an automated recommendation, accepting it without adequate scrutiny even when independent evidence should prompt a second look. Tiered decision support fights it by keeping the reversing exception and the evidence visible, but a too-confident summary tier can also feed it. ↩