Analysis · Resource
How Relationship Mapping Supports Investigative Research
How relationship mapping and link analysis clarify people, places, and events — and how to distinguish documented connections from inference.
By Second Trace editorial · Published August 30, 2026 · Updated August 30, 2026 · 6 min read
In short
Relationship mapping turns a list of names into a network of documented connections — who knows whom, who was where, what links to what — so structure becomes visible that a list hides. The discipline is to distinguish a documented connection from an inferred one, and to never let a diagram imply a link the evidence does not support.
What relationship mapping is
A relationship map is a graph: nodes represent entities — people, places, organizations, events, objects — and edges represent connections between them, each with a type and a source. Where a list forces you to read linearly, a graph lets you see clusters, bridges, and isolated facts at a glance.
Why it helps
- Pattern recognition — clusters and repeated connections become visible.
- Gap detection — a person who should be connected but is not stands out.
- Structure without judgment — the map describes, it does not accuse.
The map's value is in generating questions, not in producing conclusions. It tells you where to look, never who is guilty.
Documented vs. inferred edges
The single most important discipline in relationship mapping is labeling every edge by how it is established:
- Documented — the connection is stated or evidenced by a source you can cite.
- Inferred — you are hypothesizing a connection that the evidence does not directly show; it must carry a plain-language explanation and be marked as unverified.
Never let an inferred edge render as if it were documented. A diagram with unlabeled inference is a misinformation machine.
What to map
Map the entities that actually matter to your questions: people, places, organizations, events, and the objects that connect them (a phone, a document, a vehicle). Give each entity a stable name and note what kind of thing it is, so the map stays readable as it grows.
How to read a map without over-reading
Two nodes being connected means only that a connection exists — not its direction, not its significance, and certainly not anyone's culpability. Correlation is not causation, and centrality in a graph is not evidence of wrongdoing. Treat every surprising pattern as a question to verify, not an answer.
When to stop
A relationship map is a means to an end. Stop expanding it when it stops generating useful questions, and resist the pull to add entities simply because you can. A smaller, well-sourced map is worth far more than a large, speculative one.
Sources & further reading
- National Institute of Justice (NIJ) — research on investigative analysis and analytic methods used in serious-case review.
Frequently asked questions
What is the difference between link analysis and relationship mapping?
The terms are often used interchangeably. Both represent entities (people, places, organizations, events, objects) and the connections between them. Relationship mapping emphasizes documenting the nature and source of each connection, rather than just drawing lines.
Can a relationship map imply guilt?
It must not. A map shows that two people are connected, not that either did anything. Presumption of innocence applies to the diagram as much as to the prose: connections describe associations, not culpability, and the map should never rank or score anyone as a suspect.
How many entities before a map is useful?
Even a handful — say ten people, a few places, and a few events — can reveal structure that a flat list obscures. The value is in seeing relationships and gaps, not in having a large graph.
Put this into practice
Second Trace is a workspace for organizing case information with full provenance. These capabilities map directly to the method above:
Related resources
Timelines
Cold-case timeline →A step-by-step method for building an accurate cold-case timeline — separating exact, approximate, and disputed dates, and keeping every entry tied to its source.
Method
Facts, allegations & theories →A working taxonomy for sorting confirmed facts, allegations, testimony, inference, and theory in a case record — and why the distinction must survive export.
Reference
Glossary →Plain-language definitions of the terms used in cold-case research — from chain of custody to statute of limitations.