Reports must state boundary conditions

What it is

Boundary conditions are not a humble closing remark. They are part of the claim. Stated boundary conditions tell the reader among which people, tasks, devices, periods, and arrangements the conclusion is supported, and where the researcher explicitly does not know whether it still holds. Left unstated, readers guess—usually by inflating the finding into an unconditional truth, or by discarding the whole report as too uncertain.

Why it happens

Later users rarely reread the methods appendix. Decision makers see short sentences on the findings page; if those sentences carry no range, range is stripped away in oral retelling. Writing the boundary turns “where this does not travel” into citable text, so misuse can be pointed at. It also forces the author to separate the range already tested from the range guessed from mechanism, and to stop writing those two strengths in the same voice.

Studying it

Give each major claim a fixed set of fields: sample composition, tasks and devices, key contextual premises, uncovered groups or channels, and what extra evidence the author thinks a transfer would need. Ask a colleague who did not analyze the data to read only the findings section and mark sentences they take as established that the fields show were never covered. Compare meeting notes with the report: restrictions added in speech but missing from the document count as lost boundaries. A boundary should be checkable—“first-time older users were not included”—not “may not apply to everyone.”

Where it stops holding

A list of limits does not repair weak sampling or vague analysis. A wall of disclaimers can also drown the one limit that matters. Some constraints are ethical or confidential and cannot be expanded in an external draft; the internal draft should keep the full fields. Exploratory material may say “not yet known” over a wider area, but it still has to name what is unknown.

Applying it

• Under each finding, spend two or three sentences on for whom it holds, under what arrangement, and what was explicitly untested.
• List “we do not know” separately; do not write guesses in the same voice as observed findings.
• If a slide deck cuts methods, keep a spoken range line and keep it consistent with the document.
• In review, ask whether the written boundary tells you whom to recruit next; if it does not, it is still boilerplate.

Related

• Same group: Q4.10.1 A conclusion’s scope is bounded by the sample · Q4.10.2 A change of context can void a conclusion · Q4.10.4 Qualitative work seeks analytic generalization, not statistical representativeness · Q4.10.5 A single-case finding is not a general law until replicated · Q4.10.6 Recheck premises before transferring across cultures or platforms · Q4.10.7 Packaging small-sample findings as universal truths is a common overreach
• Adjacent: Q4.17 Communicating and acting on findings · Q1.04 Sampling and representativeness
• Search terms: boundary conditions · transferability statement · scope of claims