Within-Unit Change Assay¶
Measurement assay — instantiates Selection–Transmission Change Attribution
Measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely.
To learn whether the units themselves changed, stop averaging the population and start pairing the survivors. Within-Unit Change Assay restricts attention to units present in both states, pairs each one's before value with its after value, and averages the differences to read the transmission channel directly. Its defining discipline is that it deliberately ignores composition: entrants and exiters are excluded by construction, so no reweighting of the population can leak into the result. It is a matched, paired measurement — the transmission analogue of a within-subject study — and it answers exactly one question: among the units that stayed, did their values move, and by how much?
Example¶
A diabetes clinic's average HbA1c across its patient panel fell from 7.8% to 7.5% over a year, and the medical director wants to know whether care actually improved or whether the worst-controlled patients simply left the practice. The Within-Unit Change Assay ignores the panel average entirely. It takes only the roughly 1,400 patients enrolled in both the baseline and the follow-up window, pairs each patient's own two HbA1c readings, and averages the per-patient change: continuing patients improved by 0.2 percentage points on average. That number is the transmission term — genuine within-patient improvement among people who stayed — cleanly separated from any effect of high-A1c patients churning out of the panel. If the panel-wide drop was 0.3 points but the matched drop is only 0.2, the remaining 0.1 belongs to composition, not care.
How it works¶
Take the continuing set defined by the correspondence map; for each unit compute the difference between its later and earlier value; weight those differences by the unit's representation and average them. That weighted mean of within-unit deltas is the transmission term. Because every comparison is a unit against itself, cross-unit variation cancels and the estimate is more precise than any population contrast — the same reason paired designs beat unpaired ones. The assay's one structural blind spot is that it says nothing about the units that came or went; it trades total coverage for a clean, confound-free read of the continuing core.
Tuning parameters¶
- Continuing-set rule — strict presence in both states versus descendant-linked continuity (a unit that split still "continues" through its parts). Looser rules recover more units but lean harder on the lineage claim.
- Delta weighting — whether each within-unit change is weighted by its baseline weight, its endpoint weight, or an average. The choice shifts whether growing or shrinking units dominate the transmission estimate.
- Missing-value handling — dropping units with a missing reading versus imputing. Dropping is honest but can re-introduce a selection bias into the "continuing" set.
- Pairing granularity — individual units versus matched cohorts. Cohort pairing is robust to unit-level noise but can hide heterogeneous within-cohort movement.
When it helps, and when it misleads¶
Its strength is a clean, high-precision read on the transmission channel with the composition confound designed out: because it compares each unit to itself, it cannot be fooled by the mix changing.
Its failure mode is regression to the mean masquerading as transmission.[n1] If the continuing units were selected on an extreme baseline value — the sickest patients, the worst-performing stores — their natural drift back toward the average will register as within-unit improvement that no intervention caused. Its second limit is structural: it is silent on composition, so a flat transmission term paired with heavy churn still leaves the aggregate movement unexplained, and treating "the units barely changed" as "nothing happened" is the classic misuse. The guarding discipline is to check whether the continuing set was selected on baseline extremes, and to always report the transmission term alongside a selection measurement rather than as a standalone verdict.
How it implements the components¶
unit_value_measure— the paired before-and-after value it reads for each continuing unit.transmission_term_definition— the weighted mean of within-unit deltas is the transmission term, measured directly.state_pair_or_transition_window— it fixes the two dated states across which each unit's value is paired.
It does not itself build the identity links it relies on (unit_correspondence_map — that is the Lineage or Panel Correspondence Matrix, which it consumes) and it computes no reweighting term (selection_term_definition). Its mirror twin is the Covariance Selection-Term Calculation: this assay measures only the transmission channel over continuing units, while the covariance measures only the selection channel — the component that separates them is selection_term_definition versus transmission_term_definition.
Related¶
- Instantiates: Selection–Transmission Change Attribution — it supplies the transmission channel measured directly from continuing units.
- Consumes: Lineage or Panel Correspondence Matrix for the continuing set it pairs over.
- Sibling mechanisms: Price Equation Decomposition Table · Covariance Selection-Term Calculation · Lineage or Panel Correspondence Matrix · Composition-vs-Transformation Dashboard · Entry/Exit Normalization Protocol · Decomposition Residual Reconciliation Workflow · Selection–Transmission Sensitivity Analysis
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Within-Unit Change Assay operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely.
Independent corroboration: The frozen evidence defines Within-Unit Change Assay as 'Measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely', so its operative form is Analysis, Modeling & Optimization.
Nearest alternative: Assessment, Review & Assurance — Within-Unit Change Assay includes features of a bounded evaluation of existing evidence or work that produces a finding or disposition, but its defining operation is an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Statistics & Experimental Design
Origin pattern: Single lineage
Present-day reach: Universal
Rationale: Pairing each continuing unit's before and after measurements and analyzing the within-pair differences is the matched-pairs statistical design. NIST specifies that paired observations remove between-unit variation by basing inference on individual differences; the assay thereby isolates change from composition shifts.
Related originating lineages:
- Data Science & Analytics — Data science, analytics, and operational monitoring has a distinct contributing or parallel lineage for the mechanism's defining operation: measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely.
- Economics & Finance — Economics, finance, and mechanism-design practice has a distinct contributing or parallel lineage for the mechanism's defining operation: measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely.
- Medicine & Healthcare — medicine_healthcare contributes clinical medicine, public health, and recovery practice to this mechanism's defining operation—Measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely—without displacing the selected primary historical lineage.
- Organizational & Management Science — organizational_management contributes organizational design, management, and operational governance to this mechanism's defining operation—Measures the transmission channel directly by pairing each continuing unit's before and after value and averaging the within-unit change, ignoring composition entirely—without displacing the selected primary historical lineage.
- Systems Thinking & Cybernetics — Systems science's feedback, boundaries, stocks, flows, and regulation tradition supplies an independent formative lineage for the mechanism's within unit change assay logic.
Review resolution: The blind reviewers disagree on primary lineage (organizational_management versus statistics_experimental_design). Authoritative or primary research supports statistics_experimental_design as the best historical origin: Pairing each continuing unit's before and after measurements and analyzing the within-pair differences is the matched-pairs statistical design. NIST specifies that paired observations remove between-unit variation by basing inference on individual differences; the assay thereby isolates change from composition shifts. The cited NIST/SEMATECH e-Handbook, Paired Observations directly supports the mechanism's defining operation. All independently supported contributing domains are retained without an arbitrary cap. origin_mode=single_lineage records lineage, while domain_reach=universal records later applicability separately from provenance.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
Notes¶
[n1] Regression to the mean — when units are selected on an extreme measured value, their later measurements tend to fall closer to the average for purely statistical reasons. In a matched within-unit measurement this drift can look exactly like genuine within-unit change, which is why the continuing set's baseline selection must be checked. ↩