Process-Tracing Evidence Table¶
Evidence ledger — instantiates Multiple Causation and Explanatory Pluralism
Orders the case's evidence along its timeline and grades each piece by diagnostic strength, so a causal story must survive what actually happened, in sequence.
A causal story can be structurally elegant and still be false about the one case in front of you. Process-Tracing Evidence Table is the mechanism that checks it against the record. It lays the case's evidence out in temporal order and grades each piece not by whether it fits the favored story but by its diagnostic strength — how much its presence or absence would discriminate one hypothesis from a rival. A hypothesis earns belief only when the traces it predicts are actually there and the traces its rivals require are missing. Its defining move is within-case, sequence-respecting evidence testing: it is the sibling that asks not "how much could this factor matter?" but "what fingerprints would this cause have left, and are they on the table, in the right order?"
Example¶
An airliner overruns the end of a wet runway on landing. Instead of settling on the first plausible cause, investigators build a process-tracing evidence table. Down the rows runs the timeline: the weather report (runway wet), the recorded approach speed (slightly high), the touchdown point (long — well past the aiming zone), the thrust-reverser deployment time (delayed), the brake application (late). Each row is graded for what it discriminates. The long touchdown is a smoking gun for the "landed long, ran out of runway" hypothesis — its presence strongly confirms. The rival "brake system failed" hypothesis is put to a hoop test: it requires evidence of a brake fault, and the recorders show the brakes were functional, so that necessary evidence is absent and the hypothesis is eliminated.
Sequence does decisive work. The reverser delay occurred after the long touchdown, so it cannot be the initiating cause — at most it is an amplifier that shortened the remaining margin. The table converts a heap of undifferentiated facts into an ordered, weighted case for one causal path over its rivals, with the eliminated alternatives shown, not merely asserted.
How it works¶
- Lay the timeline as rows. Order every recorded fact by when it happened; later events cannot cause earlier ones, and the ordering is itself evidence.
- State the rival hypotheses and their predicted traces. For each candidate causal path, write down what evidence it should leave if true — before looking at what's present.
- Grade each piece by test type. Classify observed evidence as straw-in-the-wind (weakly suggestive), hoop (necessary, so its absence eliminates), smoking-gun (sufficient, so its presence confirms), or doubly-decisive.
- Keep the survivor. A hypothesis stands only if it passes its hoops and, ideally, has a smoking gun, while a rival's required-but-absent evidence is documented as the reason it fell.
Tuning parameters¶
- Evidence granularity — how finely the record is broken into rows. Finer rows catch subtle sequence effects but bury the signal in clerical detail.
- Test stringency — how strong a piece must be to count as a smoking gun or hoop. Loose thresholds confirm too easily; strict ones may leave every hypothesis unproven.
- Timeline resolution — minutes, hours, or epochs. Coarse resolution can invert the true order of fast-moving events and destroy the sequence logic.
- Rival breadth — how many competing hypotheses are tabled. Tracing only the favored story is the classic trap; more rivals cost effort but are the whole point.
- Source-reliability weighting — how much a shaky source's evidence is discounted. Over-trusting a weak source can hand a hypothesis a smoking gun it hasn't earned.
When it helps, and when it misleads¶
Its strength is that it holds a causal story to the actual record: temporal order stops consequences from posing as causes, and diagnostic grading separates evidence that truly discriminates from evidence that merely decorates. It is the sibling best at killing the archetype's post-hoc narrative fitting — the sophisticated story built after the outcome that was never tested against what the case would have had to contain.
Its failure mode is confirmation bias wearing procedural clothing: an analyst who tables only the favored hypothesis will find its traces everywhere and call the exercise rigor. Evidence can also be missing because unrecorded rather than because it never happened, so an over-read absence can wrongly eliminate a live hypothesis. And a single case, over-fit, can mistake a coincidence for a mechanism. The named discipline here is Van Evera's battery of tests — hoop, smoking-gun, and doubly-decisive — whose whole purpose is to force the analyst to specify in advance what would confirm or eliminate each rival.[n1] The guard is to table at least one genuine rival and record the evidence it predicts but that the case lacks.
How it implements the components¶
evidential_weighting— every row is graded by diagnostic strength (hoop, smoking-gun, doubly-decisive), so strongly and weakly supported causal claims never sit as equals.contingency_register— the table is the case's timeline: the sequence of events, the path-sensitive moments, and the junctures where a different turn was possible are recorded in order.
The table does not remove or vary a factor to see whether the outcome would still occur (counterfactual_probe_set, causal_weight_scale, the Counterfactual Sensitivity Probe — its nearest twin in the causal-testing lane, which reasons about imagined alternatives where this table weighs the actual record). It also does not inventory factors by family (causal_family_inventory, the Multicausal Factor Matrix) or draw the feedback structure (interaction_map, the Causal Loop or Influence Diagram).
Related¶
- Instantiates: Multiple Causation and Explanatory Pluralism — the table is the archetype's within-case evidence discipline against post-hoc storytelling.
- Consumes: a set of rival hypotheses — proposed elsewhere — that this table tests but does not itself generate.
- Sibling mechanisms: Multicausal Factor Matrix · Cause-Role Worksheet · Causal Loop or Influence Diagram · Counterfactual Sensitivity Probe · Cross-Disciplinary Causal Review
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Process-Tracing Evidence Table operates as a static representation, map, specification, schema, or prospective plan that externalizes information because it orders the case's evidence along its timeline and grades each piece by diagnostic strength, so a causal story must survive what actually happened, in sequence.
Independent corroboration: The frozen evidence defines Process-Tracing Evidence Table as 'Orders the case's evidence along its timeline and grades each piece by diagnostic strength, so a causal story must survive what actually happened, in sequence', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Political Science
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Specialized
Rationale: Process-Tracing Evidence Table is most plausibly rooted in the political_science tradition because its characteristic form depends on power, institutions, public decisions, and causal explanation across levels. The assignment tracks that formative lineage, not the many settings in which the mechanism can now be applied.
Related originating lineages:
- History & Historiography — The history_historiography tradition materially shaped Process-Tracing Evidence Table through its own practice of historical contextualization, chronology, and disciplined causal reconstruction.
- Statistics & Experimental Design — Likelihood and evidentiary-strength reasoning contributes disciplined comparison of tests for rival explanations.
Review resolution: Both blind reviewers agree that political science is the primary origin. Explicit reconciliation resolves alternate origin disagreement, encyclopedia synthesis disagreement. Formative alternate lineages are retained as history_historiography, statistics_experimental_design; later breadth of use is recorded separately as domain_reach=specialized, while origin_mode=cross_disciplinary_synthesis describes the relationship among origin lineages.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Van Evera's process-tracing tests classify evidence by what it discriminates: a hoop test supplies necessary-but-not-sufficient evidence (failing it eliminates a hypothesis), a smoking-gun test supplies sufficient-but-not-necessary evidence (passing it strongly confirms), and a doubly-decisive test does both. The tests force an analyst to say in advance what would confirm or refute each rival. ↩