Testing the evidence and its connections

A defensible finding requires a traceable path from the underlying research to the conclusion. We examine the evidence in depth and make each step of the reasoning visible enough to question, reproduce and revise.

Examine the source before relying on it

We identify who created a record, when it was created, its purpose and how it entered the research file. We examine the complete document or exchange, relevant attachments, revisions and available metadata. For quantitative material, we check definitions, units, time periods, collection methods and the calculations supporting a comparison.

We distinguish original evidence from later descriptions of it. Several reports repeating the same statement may trace back to one source. We assess the source’s opportunity to know, the completeness of the record and whether independent material supports or challenges the account. When authenticity or completeness remains uncertain, that uncertainty limits what we can conclude.

Document the connection, not just the citation

For each material claim, we record the supporting passage or data, its location, and the reasoning that connects it to the claim. A citation lets a reader find the source. The explanation must show why that source matters and exactly what it establishes.

We distinguish a documented fact from an inference, and an inference from a causal explanation. Events occurring close together may justify further inquiry. A conclusion that one caused the other requires evidence of the mechanism and consideration of other explanations. We identify which links are directly documented, which are inferred and which remain unverified.

For example, evidence that a submission was delivered does not by itself show what a reviewer considered. We examine the submission’s contents, receipt records, assignments, later acknowledgments and the reasons given for the decision. Each source answers a different part of the question.

Test the strongest competing explanation

We look for evidence that could weaken our preferred interpretation. We compare plausible explanations against the same chronology and sources, including ordinary administrative error, different definitions, incomplete production or changed circumstances where the record makes those relevant. Contrary material belongs in the account even when it makes the conclusion less straightforward.

An absent document needs particular care. We ask whether the record should exist, where it would ordinarily be kept, what was searched and whether the available production is complete. An unsuccessful search or an incomplete file can establish a limit on our knowledge without establishing that an event never occurred.

Keep interpretation within the documented context

When case law is relevant to a case study we are working on, we discuss it within that study’s factual and procedural context. We identify the specific question it helps answer, explain the connection and link the discussion to the study. Legal authority must support the issue being examined rather than serve as a detached commentary topic.

We use quotations where their actual wording matters and explain who spoke, in what role and in response to what information. Publication protects private citizens’ identities and contact information while preserving the substance needed to assess institutional conduct.

State the finding and its limits

We explain what the combined record supports, the counterevidence considered, and what remains unresolved. We check whether the conclusion extends beyond the people, period or circumstances actually examined. Where a missing link is material, we narrow the finding or return to research.

The result is an evidence-supported account that can be checked: findings, source locations, reasoning, alternative explanations and limits. That account becomes the basis for evaluating possible responses.