← Every exhibit

Table Figure 1 From the research bench

The synthetic interference exposure ledger

Before reading a unit-level effect, declare whether another unit's assignment can enter its outcome and which estimand matches that exposure structure.

IDUnit and groupOwn assignmentOther-unit assignmentExposure ruleOutcome windowEstimand
I-01Account A in buying group G1TreatedNo other treatmentOwn message delivered; no group exposure14 daysDirect effect
I-02Account B in buying group G1UntreatedAccount A treatedShared decision-maker receives message14 daysIndirect effect
I-03Account C in territory T1TreatedNeighboring accounts treatedTerritory saturation above declared threshold30 daysTotal effect
I-04Account D in territory T2UntreatedNo neighboring treatmentNo exposure map condition met30 daysComparison under design
I-05Member E in cluster K1UntreatedCluster K1 treatedCluster assignment defines exposure60 daysOverall effect under cluster assignment
I-06Account F in network N1TreatedOne linked node treatedOne-hop link exposure; link timestamp recorded21 daysDirect effect conditional on exposure

Swipe or scroll horizontally if the table is wider than your screen.

Cite Embed

Reference & Evidence

Source: Author's synthetic exposure ledger grounded in Hudgens and Halloran (2008). Assignments, exposure rules, windows, and estimands are illustrative; no commercial outcome is represented.

Each line is a claim from the register this journal publishes against, resolved from the register at build time.

  • A "a fundamental assumption usually made in causal inference is that of no interference between individuals or units that is the potential outcomes of one individual are assumed to be unaffected by the treatment assignment of other individuals" Hudgens & Halloran. (2008). Toward Causal Inference With Interference captured 2026-09-06 · HH08-C1
  • A "we propose estimands for direct indirect total and overall causal effects of treatment strategies in this setting", with relations among them established Hudgens & Halloran. (2008). Toward Causal Inference With Interference captured 2026-09-06 · HH08-C3
  • A The estimand follows the design: the paper considers "a population of groups of individuals where interference is possible between individuals within the same group" and proposes estimands for that setting Hudgens & Halloran. (2008). Toward Causal Inference With Interference captured 2026-09-06 · HH08-C4
  • A Interference changes what the outcomes ARE, not merely their direction: "the potential outcomes of one individual are assumed to be unaffected by the treatment assignment of other individuals", and where that fails the comparisons must be respecified Hudgens & Halloran. (2008). Toward Causal Inference With Interference captured 2026-09-06 · HH08-C6
  • B Interference is a situation in which one unit's outcome depends on another unit's treatment or exposure under the declared window Author framework grounded in HH08-C1 and HH08-C2 · R11-OWN-C1
  • B An exposure mapping translates own and other-unit assignments into the exposure condition used by the estimand Author framework grounded in HH08-C3 and HH08-C4 · R11-OWN-C2
  • B The six-row exposure ledger must preserve unit, own assignment, other-unit assignment, exposure rule, outcome window, and estimand Author operating framework · R11-OWN-C5
  • B The illustrative ledger contains no campaign, customer, territory, or network outcome data Author synthetic object · R11-OWN-C8

Grades: A, verified against the printed page of the primary source · B, primary source, text layer only · C, authoritative secondary · D, reported.

Related exhibits