Status: v1, 27 April 2026.

Why this document exists

Tourism research is rarely transparent about how it was actually produced. Source selection, translation choices, AI tool use, verification practices, and the gap between what was read and what was synthesised are usually invisible to the reader. That invisibility is convenient for the researcher but unhelpful for anyone who wants to assess the work seriously.

This project has decided to make those choices explicit. This document sets out, in plain language:

If a critic of this work wants to identify methodological weakness, this file should make that easy. If a stakeholder wants to assess whether the research meets a serious standard, this file should make that easy. The intent is to be auditable.

Sources

The research draws on three categories of source.

Primary documents are the strategy, policy, and planning documents produced by Finnish destinations, municipalities, and tourism bodies, plus comparable academic and policy output from international sources. The current core set is:

The full source list, including the academic papers, is published alongside this document.

Secondary research consists of the wider tourism studies literature accessed through the University of Lapland Research Portal, the Multidimensional Tourism Institute, peer-reviewed journals, and policy publications. Where this work draws on a secondary source, the citation includes a link to the original. The project does not synthesise from second-hand summaries when first-hand access is available.

Primary research consists of structured conversations with local providers, residents, academics, and stakeholders in Levi and the wider Lapland region. This work is in early stages. Where conversations are referenced, the participant is named with their consent, the date and approximate length of the conversation are recorded, and the citation makes clear what was said versus what is being inferred from what was said.

Sources are selected on three criteria: relevance to the research question, accessibility for verification, and the standing of the publishing body. Sources that cannot be verified by a third-party reader are not used.

The language question

The author of this work does not speak Finnish at a level required to read strategy documents in their original language without translation support. Several of the most important primary sources are published only in Finnish.

This is a real methodological constraint. Strategy documents are written in a register where tone, hedging, irony, bureaucratic euphemism, and the difference between “stated commitment” and “aspirational language” can determine the meaning of a passage. AI-assisted translation is broadly reliable for content but unreliable for register. A faithful read of these documents requires native-speaker engagement at the verification stage.

The project’s response to this is structural:

The longer-term answer to the language gap is a Finnish-language collaborator working as a research partner, not as a translator. The project considers itself unfinished until that relationship is in place.

AI-assisted research

This project uses AI tools substantially. Pretending otherwise would be dishonest. The honest approach is to name what is used, for what, and what the limits are.

Tools used

The project has used, and continues to use:

Each of these tools has been used for a different role. Each has known limitations. None of them is treated as a primary research output. AI-produced material is treated as a hypothesis until verified against the underlying source.

What AI is used for

What AI is not used for

Known limitations of the tools used

The verification log accompanying this document records what was actually found when AI-produced citations were checked against the original sources. Headline findings:

Verification protocol

In response to these findings, the project has adopted the following verification protocol. Any citation appearing in published material on this site must satisfy all of the following before publication:

  1. Page reference verified against the original document. No published page number is taken from AI synthesis without direct check.
  2. Quote text verified against the original document. Quotations are reproduced from the source, not from AI output.
  3. Surrounding context reviewed. The passage’s position in the document — strategic commitment vs workshop output, body text vs sticky-note brainstorm, current vs historical — is established before the framing is fixed.
  4. For Finnish-language passages, native-speaker verification of translation and register. The Finnish original, the English translation, and the framing are reviewed by a Finnish-language reader.
  5. For inferred or interpretive claims, explicit labelling. Where a claim is the project’s analytical interpretation rather than a direct document statement, the text says so. “The strategy implies” is different from “the strategy commits to”. The difference is preserved.

Citations that fail any of these checks are either corrected, downgraded to “unverified”, or removed.

Disclosure in published material

Where AI tools have shaped published material substantively, the methodology is disclosed. Where AI has been used for routine drafting and editing, this document serves as the standing disclosure rather than per-page attribution. Readers can assume that any text on this site has been drafted and edited with AI assistance unless attributed otherwise. They can also assume that the substance — the arguments, the claims, the verification — is the responsibility of the named author.

This is the standard the project holds itself to. AI is a research collaborator that cannot be cited and cannot be held accountable. The buck stops with the human author.

Verification log

A continuously-maintained verification log records every citation checked against its original source, what was found, and where corrections were applied. The log is published as a separate document and is updated as further verification is completed.

The log currently covers six core claims about Levi/Lapland tourism strategy across the four primary Finnish-language documents and the IPOL European Parliament study. Approximately forty further citations across the secondary academic literature and topic quote banks have not yet been verified. The log is explicit about what is verified and what is not.

The log itself is part of the project’s methodology, not a separate accounting exercise. Its existence is the point: the work has been checked, the checking is open, and where the checking found errors, the errors are recorded with the correction.

Corrections

Substantive errors in the research are corrected publicly. The correction policy is:

Silent corrections are not made. Removing an error without acknowledgement is itself a form of error.

What this methodology cannot do

In the interest of not overclaiming, the following limitations are explicit:

How to challenge this methodology

If you believe the methodology described here is insufficient, biased, or wrong, the project would like to know:

Substantive challenges are published, with permission, alongside the project’s response. The methodology is itself version 1.0 and is open to revision.

Author: Colin Harrison, Levi, Finland. Contact: colin@levifinland.com This file is part of the methodology layer of future.levifinland.com. See also: Ethics and Positionality, Verification Log.