Triangulation Matrix¶
Template — instantiates Source Provenance Triangulation
Places claims against multiple sources so agreement, disagreement, independence, and source-type diversity can be seen at once.
A list of sources tells you how many you have; it does not tell you which claims they actually cover, or whether the ones that agree are truly independent. Triangulation Matrix is the template that turns evidence into a grid — claims down one axis, sources across the other — so that for every claim you can see at a glance who supports it, who is silent, who contradicts, and whether the supporters are diverse and independent or all of one kind. Its defining move is cross-tabulation for convergence: by laying claims against sources it exposes where agreement is real (independent sources of different types converging) and where it is illusory (a claim resting on one source, or on several copies of one origin), and it lets those columns be weighted rather than merely counted. It is built to find and strengthen corroboration, not to preserve disagreement.
Example¶
A public-health team is reconstructing the origin of a foodborne outbreak. They have five source streams: patient case reports, laboratory pathogen typing, restaurant purchase records, contact-tracing interviews, and a supplier's distribution log. Rather than narrate from the most alarming interview, they build a triangulation matrix. Each candidate claim — "the implicated ingredient is batch X," "onset clusters on the 12th," "supplier S is common to all cases" — becomes a row; each source stream a column.
The grid immediately does work. The claim that supplier S is common is supported by purchase records, the distribution log, and lab typing — three sources of different types (documentary, transactional, biological) that don't share an origin, so their agreement is strong corroboration. The claim about batch X, by contrast, lights up in only one column: a single interview. And two "independent" case reports turn out to trace to the same clinic's intake form — copies, not corroboration, so they are down-weighted to count as one. Seen as a grid, the well-supported claims and the singly-sourced ones separate cleanly, and the team can weight the convergent, diverse evidence over the lone vivid account.
How it works¶
- Build the grid. Claims as rows, sources as columns; each cell records whether the source supports, contradicts, is silent on, or is irrelevant to the claim.
- Read down each column for type. Tag each source by kind (documentary, testimonial, instrumental, biological) so a row's support can be judged for diversity, not just count.
- Read across each row for convergence. A claim backed by several independent, differently-typed sources is corroborated; a claim in one cell, or in cells that share an origin, is not.
- Collapse copies and weight the rest. Merge sources that trace to one origin so they count once, then weight surviving support by independence and type diversity rather than tallying cells.
Tuning parameters¶
- Cell vocabulary — binary supports/silent, or a richer set (supports, contradicts, partially supports, silent, irrelevant). Richer cells carry more but demand more adjudication per claim.
- Independence collapsing — how aggressively you merge sources that share an origin before counting support. Aggressive collapsing kills false corroboration but can under-count genuinely semi-independent sources.
- Type-diversity weighting — how much extra weight convergence across different source types earns over agreement within one type. High weighting rewards true triangulation; too high penalizes domains where one source type dominates.
- Threshold for "corroborated" — how many independent, diverse sources a row needs before it is treated as well-supported. A high bar resists premature confidence but leaves more claims provisional.
When it helps, and when it misleads¶
Its strength is that it makes the structure of support visible: the singly-sourced claim can no longer hide inside a coherent story, and agreement is credited only when the agreeing sources are independent and diverse. This is the working form of methodological triangulation — combining differently-vulnerable sources so their convergence means more than any one could alone.[n1] It is the backbone of incident postmortems, mixed-method research, and outbreak reconstruction.
Its failure mode is illusory corroboration: sources can share a hidden dependency the grid doesn't capture — a common upstream feed, a shared assumption — so cells that look independent are secretly one. It can also flatten a genuine, meaningful contradiction into a mere "disagreement" cell and move on, losing the signal that a conflict carries. The classic misuse is counting cells instead of weighing them — treating five supporting sources as five times the confidence without checking whether they are independent or of different types. The discipline that guards it is to collapse shared origins before counting and to hand rows that stay genuinely contradictory to a format built to hold them open rather than resolving them by majority.
How it implements the components¶
corroboration_check— reading across a row tests whether agreeing sources are independent and differently vulnerable, so convergence counts only when it is real.source_type_classification— tagging each column by source kind makes a row's diversity legible, distinguishing broad triangulation from agreement within a single type.reliability_weighting_rule— support is weighted by independence and type diversity, and copies are collapsed to count once, rather than claims being settled by a raw tally of cells.
It cross-tabs to reveal and strengthen agreement, so it does not keep an unresolved contradiction standing as its own record, register the open question that would settle it, or attach a contested per-claim label (dissent_record, source_gap_register, confidence_label — that is Conflicting Source Table, its nearest twin, which deliberately holds disagreement open where this matrix works to resolve it).
Related¶
- Instantiates: Source Provenance Triangulation — supplies the claim-by-source grid that separates genuinely corroborated claims from singly-sourced ones.
- Consumes: Citation Lineage Review — supplies the independence findings that tell the matrix which apparently separate sources are actually copies to collapse.
- Sibling mechanisms: Conflicting Source Table · Audit Trail Review · Citation Lineage Review · Source Criticism Protocol · Witness / Source Comparison · Chain-of-Custody Record
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Triangulation Matrix is defined in the frozen evidence as: Places claims against multiple sources so agreement, disagreement, independence, and source-type diversity can be seen at once. Its operative deployed or enacted form is therefore Representation, Specification & Plan.
Nearest alternative: Interface, Display & Cue — Interface, Display & Cue can support this mechanism, but the evidence centers the concrete operation described above rather than the alternative family's defining operation.
Review outcome: Adjudicated after independent review; medium confidence.
Origin Attribution¶
Primary origin: Ethnography & Qualitative Methods
Origin pattern: Convergent development
Present-day reach: Universal
Rationale: A claims-by-sources matrix operationalizes methodological triangulation: each conclusion is checked against evidence produced through more than one source or method. CDC explicitly defines triangulation as cross-verification using multiple sources and methods; statistics supplies formal comparison.
Related originating lineages:
- Accounting & Auditing — accounting_auditing contributes accounting, auditing, and controlled-resource stewardship to this mechanism's defining operation—Places claims against multiple sources so agreement, disagreement, independence, and source-type diversity can be seen at once—without displacing the selected primary historical lineage.
- Data Science & Analytics — Data modeling, telemetry, and analytic monitoring supplies a distinct formative lineage for the mechanism's triangulation matrix logic.
- History & Historiography — Historical and historiographic method supplies a parallel or contributing lineage for the mechanism's defining operation: places claims against multiple sources so agreement, disagreement, independence, and source-type diversity can be seen at once.
- Sociology & Anthropology — Sociology and anthropological study of institutions and social relations supplies a parallel or contributing lineage for the mechanism's defining operation: places claims against multiple sources so agreement, disagreement, independence, and source-type diversity can be seen at once.
- Statistics & Experimental Design — Statistics, experimental design, and measurement theory supplies a parallel or contributing lineage for the mechanism's defining operation: places claims against multiple sources so agreement, disagreement, independence, and source-type diversity can be seen at once.
Review resolution: The blind reviewers disagree on primary lineage (statistics_experimental_design versus ethnography_qualitative_methods). Authoritative or primary research supports ethnography_qualitative_methods as the best historical origin: A claims-by-sources matrix operationalizes methodological triangulation: each conclusion is checked against evidence produced through more than one source or method. CDC explicitly defines triangulation as cross-verification using multiple sources and methods; statistics supplies formal comparison. The cited CDC, Gathering Credible Evidence and Triangulation; HM Treasury, The Magenta Book directly supports the mechanism's defining operation. All independently supported contributing domains are retained without an arbitrary cap. origin_mode=convergent records lineage, while domain_reach=universal records later applicability separately from provenance.
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:
Notes¶
[n1] Methodological triangulation — a term borrowed from surveying and navigation and adapted in the social sciences (notably by Norman Denzin) — is the practice of establishing a finding through multiple, differently-vulnerable sources or methods, on the logic that convergence from independent vantage points is stronger evidence than depth from any single one. The matrix is the tabular instrument that makes such convergence, and its absence, visible. ↩