Skip to content
Prototype data

Records marked Prototype are illustrative examples demonstrating the platform data model — not verified Jarlhalla research, and not citable as findings. The absence of that marker is not itself evidence: treat any record as unverified until it shows its sources and a verification date, and note that a workflow status is never proof that a claim is true. Read the provenance note.

About & Methodology

How we separate what we know from what we think

The hub is a working research register, not a publication archive. Records are visible while they are still uncertain — which only works if the difference between a source, an observation, an analysis, a hypothesis and a finding is impossible to miss.

The chain

Source → observation → analysis → hypothesis → finding

  1. 01

    Source

    An external artefact: a paper, standard, dataset, regulation, report or interview. A source is cited, never paraphrased into a claim of our own. It carries a type, a URL or reference, a publication date and an ingestion date.

  2. 02

    Observation

    Something recorded directly — a field measurement, a document change, a market movement. An observation states what was seen, when, where and by which method. It is not yet an explanation.

  3. 03

    Analysis

    Structured reasoning over sources and observations. Analysis is interpretive by definition and is labelled as such throughout the platform. It can be wrong while every input beneath it is correct.

  4. 04

    Hypothesis

    A proposition we consider plausible and intend to test. Hypotheses are never displayed as findings. Records of this type carry an explicit interpretive warning on their page.

  5. 05

    Finding

    A conclusion we are willing to stand behind, with its sources attached, an evidence level assigned and a verification date. A finding can still be revised — and its revision history stays visible when it is.

Evidence

Evidence levels

Evidence level describes the strength of the underlying material. It is independent of how far a record has travelled through our workflow.

  • Anecdotal

    A single uncorroborated account or isolated report. Useful as a lead; never sufficient on its own.

    Anecdotal

  • Preliminary

    Early data or a first result, not yet replicated or independently checked.

    Preliminary

  • Corroborated

    Multiple independent sources point the same way, or a result has been reproduced.

    Corroborated

  • Peer reviewed

    Published in a peer-reviewed venue, or reviewed to an equivalent external standard.

    Peer reviewed

  • Authoritative

    Issued by the body that defines the matter — a standards organisation, regulator or official statistics provider.

    Authoritative

Workflow

Verification statuses

Status says where a record is in our process. A verified status means we have checked our own chain of reasoning and sources — it is not a guarantee about the external world.

Verification statuses used across the platform
StatusWhat it means
Workflow status: DraftBeing written. Incomplete and not reviewed. Visible for transparency, not for citation.
Workflow status: MonitoringWe are watching this and collecting material. No conclusion is claimed.
Workflow status: Active researchUnder active investigation with allocated effort and a defined question.
Workflow status: Peer reviewComplete and under internal or external review before promotion.
Workflow status: VerifiedSources checked, reasoning reviewed, verification date recorded. Still revisable.
Workflow status: ArchivedRetained for the record but no longer maintained or re-verified.
Workflow status: RetractedWithdrawn. Kept visible so the correction is part of the public record.

Principles

Editorial and research principles

  • Traceability

    Every analytical record links to the sources and observations it rests on. If a claim has no traceable input, it does not get published.

  • Corrections in the open

    Errors are corrected visibly. We amend the record and note that it was amended, rather than silently replacing text.

  • Versioning

    Publications and research records carry revision numbers. A citation points at a specific revision, not a moving target.

  • Source transparency

    Source type, publisher, jurisdiction and access date are shown. Where a source is preliminary, a preprint or an internal note, that is stated.

  • Uncertainty is stated

    Confidence levels and evidence levels are part of the record, not a footnote. Low confidence is published as low confidence.

  • No hypothesis presented as fact

    Workflow status describes where a record sits in our process. It is not a truth claim about the world. Nothing marked hypothesis, analysis, scenario or risk brief should be read as an established result.

Provenance

On the records currently in this platform

Most records visible here are prototype data, and each one carries a Prototype marker. They exist to demonstrate the data model — statuses, evidence levels, source links, revisions, topic aggregation — and to show how the platform behaves with content in it. They are illustrative examples, not verified Jarlhalla research, and at least one reference was a placeholder rather than a real identifier.

Records without that marker are real Jarlhalla records. Real does not mean proven: a record only carries factual claims once sources and evidence are attached to it, and a workflow status describes editorial progress rather than the truth of a claim.

Nothing marked as prototype should be cited, quoted or treated as a factual claim about the world. When verified records replace them, the prototype notice across the platform will be removed.

Ecosystem

Part of Jarlhalla

The Research Hub is an independent daughter platform of Jarlhalla. It runs on its own data model, its own editorial process and its own hosting, so that it can be moved to a dedicated address such as research.jarlhalla.no without touching the parent site. Its findings inform Jarlhalla's work; Jarlhalla's commercial activity does not set its conclusions.

jarlhalla.no