A number without lineage is a claim, not evidence
"Evidence-based" has become a phrase institutions attach to almost any report that includes a chart. The phrase does real work only when it points to something specific and checkable. Evidence lineage is that specific thing: the ability to trace any number on a dashboard back to its source data, the method used to collect it, and every transformation applied along the way.
Most dashboards fail this test the moment someone asks a direct question about them. A graph appears on screen. No one in the room can say where the underlying data came from, how it was gathered, or what happened to it between collection and the chart. That gap is not a cosmetic problem. It is the difference between an institution that can defend a decision under scrutiny and one that is guessing with confidence.
The four links in the chain
Lineage has four parts, and all four have to hold for a number to count as evidence rather than decoration.
Source. Where did the raw data originate — a sensor reading, a survey response, a transaction record, an administrative filing? A number without an identifiable source is not traceable by definition.
Collection method. How was it gathered, by whom, and under what conditions? A figure collected by a trained enumerator following a documented protocol carries different weight than one pulled from an unverified third-party feed, even if both produce the same number on a dashboard.
Transformation. What happened to the number between collection and display? Was it aggregated, normalized, adjusted for inflation, deduplicated, imputed where missing? Every transformation is a place where an error can enter or a bias can be introduced, and each one needs to be documented, not assumed.
Audit trail. Can someone who was not in the room when the number was produced reconstruct how it got there, using only the documentation? If reconstruction requires calling the one analyst who remembers, the lineage exists in a person's memory, not in the institution's evidence base — and that person eventually leaves.
Why this matters more than data volume
Institutions under pressure to modernize often respond by collecting more data — more sensors, more surveys, more integrations. Evidence quality is a separate variable from data volume, and it is the one that actually determines whether a decision can survive an audit. Ten data points with full lineage produce a more defensible decision than ten thousand data points with none, because the ten thousand cannot answer a board member's basic question: how do we know this is right.
The test that actually matters
The practical test for evidence lineage is simple to state and uncomfortable to apply: pick any number on any dashboard currently informing an institutional decision, and ask someone outside the team that built it to trace it back to source, method, and transformation, using only the documentation available. If they cannot, the number is not evidence. It is a claim wearing the visual language of evidence — and until an institution can consistently pass this test, more dashboards will not fix the underlying problem.