TOPICALAUTHORITY.ORG TAO / ROOT

Reproducible Digital Research

TOPICALAUTHORITY.ORG / METHODS / MTH-10STATE REPRODUCTION CONTROL ACTIVE
MTH / 10 · VERSION · SNAPSHOT · LINEAGE · RERUN · DRIFT

Reproducible Digital Research.

Reproducibility is not getting the same number forever. It is preserving enough of the research system to rebuild a result, determine why a rerun agrees or differs, and separate methodological failure from genuine change in the digital environment.

01 / VERSIONWhich method?
02 / FREEZEWhich inputs?
03 / TRACEWhich transformations?
04 / EXECUTEWhich decisions?
05 / COMPAREWhat changed?
06 / EXPLAINWhy did it change?
MTH / 10.1 OPERATIONAL DEFINITION

A result is reproducible when its production can be reconstructed.

Dynamic digital environments may change between runs. A valid reproduction therefore preserves the method while explicitly modeling source drift, access loss and time-dependent state.

TAO / DEFINITION
Reproducible digital research is the controlled preservation and re-execution of a versioned research process—including questions, boundaries, inputs, snapshots, transformations, decision rules and outputs—so agreement and difference between runs can be independently traced and explained.
REPRODUCIBLE ≠ PERMANENTLY IDENTICALThe web can change while the method remains stable.

A different output may be correct when a new snapshot contains a real state change.

REPEAT ≠ RECONSTRUCTRunning an interface twice is not enough.

Inputs, versions, filters, transformations and decisions must also be recoverable.

CODE ≠ COMPLETE METHODExecutable logic is only one layer.

Human inclusion decisions, unresolved states and interpretation rules also require records.

MTH / 10.2 REPRODUCTION CONTRACT

Nine invariants define what must be preserved.

The contract separates stable research logic from time-sensitive environmental state.

CONTROL / 01

Question version

Preserve the exact operational question, claim form and intended decision.

LOCK / SEMANTIC TARGET
CONTROL / 02

Boundary version

Record inclusion, exclusion, language, geography, time and unit definitions.

LOCK / ELIGIBLE UNIVERSE
CONTROL / 03

Source identity

Resolve the source object, access route, version and observed location.

LOCK / SOURCE REGISTRY
CONTROL / 04

Snapshot state

Freeze or fingerprint eligible content at the declared observation time.

LOCK / TIME-BOUND INPUT
CONTROL / 05

Collection rule

Preserve request parameters, pagination, ordering, retry and missing-state rules.

LOCK / CAPTURE PROCEDURE
CONTROL / 06

Transformation log

Record normalization, filtering, joining, calculation and manual revision steps.

LOCK / DATA LINEAGE
CONTROL / 07

Decision rules

Version thresholds, classifications, conflict treatment and reviewer overrides.

LOCK / JUDGMENT LAYER
CONTROL / 08

Output schema

Preserve field meanings, units, null behavior, sort order and expected artifacts.

LOCK / RESULT CONTRACT
CONTROL / 09

Execution context

Record material environmental conditions that can change process behavior.

LOCK / RUN CONTEXT
MTH / 10.3 TEN-STAGE REPRODUCTION

Freeze logic. Rebuild outputs. Explain every delta.

A reproduction run is not successful merely because it finishes. It must establish whether each divergence originates in the method, input state, execution, decision layer or digital environment.

01 / IDENTIFY

Select baseline

Resolve the exact research run and result to reproduce.

OUTPUT / RUN ID
02 / VERIFY

Inspect manifest

Confirm required components and known limitations.

OUTPUT / READINESS STATE
03 / RESTORE

Load method

Recover question, boundary and decision versions.

OUTPUT / METHOD PARITY
04 / RESOLVE

Load inputs

Use baseline snapshots or declared new-time sources.

OUTPUT / INPUT PARITY
05 / COLLECT

Repeat capture

Execute the same request, traversal and state rules.

OUTPUT / RAW RERUN
06 / TRANSFORM

Rebuild dataset

Apply the preserved sequence and calculations.

OUTPUT / DERIVED DATA
07 / DECIDE

Reapply rules

Classify observations using the same decision model.

OUTPUT / DECISION LOG
08 / COMPARE

Calculate parity

Compare inputs, intermediate states and outputs.

OUTPUT / DELTA MATRIX
09 / ATTRIBUTE

Explain drift

Assign each delta to a traceable change event.

OUTPUT / DRIFT REGISTER
10 / PUBLISH

Issue outcome

Record reproducibility state, differences and limits.

OUTPUT / RERUN RECORD
MTH / 10.4 REPRODUCTION CONSOLE

Four operations. Every protocol remains visible.

Without scripting all protocols display in sequence. With scripting they operate as a focused control console.

TAO REPRODUCTION OPERATIONS / RR-010SELECT OPERATION
OPERATION / BASELINE CAPTURE

Freeze everything needed to reconstruct the run

STATE / COMPLETE
METHODQuestion, scope, sample, validation and decision versions.
INPUTSSource identities, parameters, snapshot times and missing states.
PROCEDUREOrdered collection and transformation instructions.
OUTPUTSRaw, intermediate and final artifacts with declared schemas.
FINGERPRINTSContent-derived identifiers used to detect byte or record change.
LIMITSUnavailable components and irreducible manual decisions.
NOT PERMITTED / A final spreadsheet or published chart alone is not a reproducible research package.
OPERATION / TRANSFORMATION GRAPH

Connect every output to its contributing inputs

STATE / TRACEABLE
NODESSources, snapshots, extracted records, normalized tables, decisions and findings.
EDGESCollected from, filtered by, joined with, derived through and reviewed by.
ORDERMaterial sequence and dependency direction remain explicit.
TRANSFORMSInputs, formulas, null rules and outputs for every derivation.
MANUAL EVENTSReviewer changes are first-class lineage events, not hidden edits.
OUTPUTDirected run graph from origin snapshots to published claims.
NOT PERMITTED / A named script or tool does not replace a record of what inputs and rules it actually used.
OPERATION / CONTROLLED EXECUTION

Rebuild the result under a declared rerun mode

STATE / EXECUTABLE
MODE ASame-snapshot reproduction tests method and execution parity.
MODE BNew-snapshot reproduction tests stable method against environmental change.
ORDERExecute preserved steps without outcome-led modifications.
CHECKPOINTSCompare raw, intermediate, decision and final states.
ERRORSRecord access failure, missing input, ambiguity and execution deviation.
OUTPUTComplete rerun package linked to its baseline record.
NOT PERMITTED / Do not modify rules during the rerun to force agreement with the original result.
OPERATION / DELTA ATTRIBUTION

Explain where and why the two runs diverge

STATE / BOUNDED
INPUT DRIFTSource content, availability, ordering or response state changed.
METHOD DRIFTQuestion, boundary, collection or transformation version changed.
DECISION DRIFTClassification or reviewer judgment changed under equal evidence.
OUTPUT DRIFTSchema, sort, formatting or calculation result differs.
UNEXPLAINEDDifference remains after all available lineage checks.
OUTPUTIdentical, equivalent, explainable, unexplained or non-reproducible.
NOT PERMITTED / A changed result cannot be called a failure until method drift and real environmental change are separated.
MTH / 10.5 RUN MANIFEST

The complete run, readable as one control object.

This synthetic manifest shows the minimum relationships needed to recover a topical coverage audit without pretending that a filename alone preserves the method.

SYNTHETIC MANIFEST / TOPICAL AUDIT

One run identity. Every dependency declared.

The manifest does not need to store every source inside one object. It must resolve each required component unambiguously and preserve its version, location and role.

Secrets and access credentials are never part of the published record. Their required capability and access condition can be documented without exposing them.

RUN / TAO-COV-2026-08-01-AMANIFEST VALID
question_versionRQ-COVERAGE / v1.3
boundary_versionCORPUS-BOUNDARY / v2.1
requirement_setTOPICAL-REQ / v4.0 / 100 eligible
source_snapshotSNAP-20260801 / 124 documents
collection_ruleCOLLECT-WEB / v1.7
transform_chainextract → normalize → resolve → classify / v3.2
decision_modelCOVERAGE-STATES / v1.2
output_schemaCOV-MATRIX / v2.0 / 9 fields
resultsupported_coverage / 72.0%
known_limits4 unknown freshness states / disclosed
MTH / 10.6 LINEAGE MAP

Every finding has a path back to evidence.

The graph makes intermediate datasets and decision states visible instead of drawing a false direct line from source pages to a final percentage.

01 / SOURCE LAYER
S-001Corpus snapshot124 eligible documents
S-002Requirement universe100 eligible obligations
02 / OBSERVATION LAYER
O-001Extracted passageslocation + source ID
O-002Detected relationssubject + predicate + object
03 / DECISION LAYER
D-001Coverage statessupported / partial / absent
D-002Ownership statesprimary / support / conflict
04 / OUTPUT LAYER
R-001Coverage matrixrequirement-level results
R-002Published finding72.0% supported
TRACE RULE / The final finding is reproducible only if its contributing requirements, observations, decisions and transformations can be resolved through the lineage graph.
MTH / 10.7 DRIFT MATRIX

A changed output becomes useful when its cause is located.

The synthetic rerun uses the same topical requirement model one month later. Three source changes explain the result delta.

COMPONENTRUN ARUN BPARITYDELTA ORIGINRESULT EFFECTDECISION
Question versionv1.3v1.3IDENTICALNoneNoneMethod stable
Requirement universe100 eligible100 eligibleIDENTICALNoneDenominator stableMethod stable
Source corpus124 documents121 documentsCHANGED2 removed; 1 inaccessible5 evidence paths affectedEnvironmental drift
Transformation chainv3.2v3.2IDENTICALNoneNoneExecution stable
Coverage decisions72 supported68 supported−4 ITEMSLost valid evidence−4.0 ppExplained result drift
Freshness decisions4 unknown6 unknown+2 UNKNOWNSource dates removedUncertainty risesDisclose limitation
VALID READING / Run B does not disprove Run A. Both describe different corpus snapshots under the same method. The result is reproducible with explainable environmental drift.SYNTHETIC DATA / NOT A LIVE CLAIM
MTH / 10.8 OUTCOME TAXONOMY

Five rerun outcomes. Only one is a hard failure.

Reproducibility language must identify whether equality was expected and which components changed.

OUTCOME / 01

Identical

Declared inputs, intermediate states and final outputs match under the comparison rule.

SAME SNAPSHOT / SAME METHOD
OUTCOME / 02

Equivalent

Non-material formatting or ordering differs while analytical values and decisions match.

SEMANTIC PARITY / FORMAT DRIFT
OUTCOME / 03

Explainable drift

Output differs and every material delta maps to a documented source, method or context change.

VALID RERUN / ATTRIBUTED CHANGE
OUTCOME / 04

Unexplained drift

Execution completes but one or more material differences cannot be located in available lineage.

OPEN INVESTIGATION / WITHHOLD
OUTCOME / 05

Non-reproducible

Required inputs, rules, versions or decisions are missing, preventing controlled reconstruction.

HARD FAILURE / PACKAGE INCOMPLETE
MTH / 10.9 REPRODUCTION MEASURES

Six measures. No single reproducibility score.

Each measure diagnoses a different part of the research package or rerun result.

MEASURE / 01

Manifest completion

available required components ÷ required components

Shows whether the package contains everything its reproduction contract demands.

READINESS / COMPONENT LEVEL
MEASURE / 02

Input identity rate

resolved input objects ÷ declared input objects

Tests whether source and snapshot references identify usable objects unambiguously.

INPUT / IDENTITY LEVEL
MEASURE / 03

Lineage completion

traceable output fields ÷ tested output fields

Measures whether each tested result can be reconstructed through its dependencies.

TRACE / FIELD LEVEL
MEASURE / 04

Decision agreement

matching decisions ÷ comparable decisions

Excludes cases whose inputs legitimately changed between observation windows.

PARITY / DECISION LEVEL
MEASURE / 05

Drift attribution

explained material deltas ÷ material deltas

Shows how much rerun difference can be connected to documented change events.

DIAGNOSIS / DELTA LEVEL
MEASURE / 06

Unresolved divergence

unexplained material deltas ÷ compared outputs

Preserves differences that remain open after available method and lineage checks.

RISK / OUTPUT LEVEL
MTH / 10.10 WORKED EXAMPLES

Four reruns. Four different valid interpretations.

All values are synthetic and illustrate methodology rather than live external findings.

SAM
EXAMPLE / 01 · SAME SNAPSHOTCoverage audit rebuilt from preserved inputs
ILLUSTRATIVE
BASELINE72 supported of 100 eligible.
RERUNSame snapshot and method versions.
OUTPUT72 supported; identical requirement states.
DECISIONIdentical reproduction.
LIMIT / IDENTITY IS CLAIMED ONLY FOR THE TESTED FIELDS AND COMPARISON RULE
WEB
EXAMPLE / 02 · NEW WEB SNAPSHOTThe method matches while the corpus changes
ILLUSTRATIVE
METHODQuestion, universe and rules unchanged.
ENVIRONMENTThree pages removed; two documents updated.
OUTPUTSupported coverage moves from 72% to 68%.
DECISIONExplainable environmental drift.
LIMIT / DIFFERENT OUTPUT DOES NOT MEAN THE ORIGINAL METHOD FAILED
HUM
EXAMPLE / 03 · REVIEWER DECISIONTwo reviewers classify one ambiguous observation differently
ILLUSTRATIVE
INPUTSame passage, requirement and codebook.
RUN AReviewer assigns partial coverage.
RUN BReviewer assigns contextual mention.
DECISIONDecision drift; adjudication required.
LIMIT / HUMAN JUDGMENT MUST BE VERSIONED AND REVIEWABLE, NOT PRETENDED AWAY
MIS
EXAMPLE / 04 · MISSING TRANSFORMA published index cannot be rebuilt
ILLUSTRATIVE
AVAILABLERaw observations and final score.
MISSINGWeighting, null and denominator rules.
ATTEMPTMultiple plausible calculations produce different scores.
DECISIONNon-reproducible package.
LIMIT / A FINAL NUMBER CANNOT REPLACE ITS TRANSFORMATION LOG
MTH / 10.11 REPRODUCTION FAILURES

Eight omissions that break the reconstruction path.

These failures may leave a polished final report while making its production impossible to inspect or repeat.

FAILURE / 01

Unversioned question

The analytical target changes without a new method identity.

CONTROL / VERSION SEMANTICS
FAILURE / 02

Live-source substitution

A current page silently replaces the historical state used by the baseline.

CONTROL / FREEZE SNAPSHOT
FAILURE / 03

Parameter loss

Location, language, ordering, pagination or filtering inputs are unavailable.

CONTROL / LOG REQUEST STATE
FAILURE / 04

Hidden manual edit

A reviewer changes records outside the documented transformation chain.

CONTROL / RECORD EVENTS
FAILURE / 05

Schema drift

A field changes meaning or null behavior without an output version change.

CONTROL / VERSION CONTRACT
FAILURE / 06

Outcome forcing

Rules are adjusted during rerun to reproduce the expected number.

CONTROL / FREEZE DECISIONS
FAILURE / 07

Secret dependency

An undocumented private input materially influences the result.

CONTROL / DECLARE CAPABILITY
FAILURE / 08

Drift denial

A genuine digital state change is mislabeled as methodological inconsistency.

CONTROL / ATTRIBUTE DELTAS
MTH / 10.12 REPRODUCTION RECORD

Every rerun should be auditable as its own research object.

The record links baseline and reproduction without overwriting either state.

REPRODUCTION RECORD / MTH10-RERUN-V1
01 / RUN IDENTITYBaseline and rerun IDsparent + mode + execution time
02 / METHODVersioned research systemquestion + boundary + sample + rules
03 / INPUTSSource and snapshot registryidentity + state + fingerprint
04 / LINEAGETransformation graphsteps + dependencies + manual events
05 / OUTPUTSIntermediate and final artifactsschema + field + comparison rule
06 / PARITYComponent comparison statesidentical + equivalent + changed
07 / DRIFTMaterial delta registerorigin + impact + attribution
08 / LIMITSUnavailable or uncertain elementsmissing + inaccessible + unresolved
09 / OUTCOMEBounded reproduction decisionstate + reason + reviewer + version
MTH / 10 · REPRODUCTION PRINCIPLE

The result may change. The reasoning path must remain visible.

Reproducible digital research preserves the system that produced a finding, rebuilds it under declared conditions and turns every material difference into an inspectable research event.

TOPICALAUTHORITY.ORG / METHODS / MTH-10VERSION → FREEZE → TRACE → REBUILD → COMPARE → EXPLAIN → RECORD
TAO / CONTACT · DIRECT TRANSMISSION Have an asset, domain or market position to investigate? ENTER CONTACT SYSTEM →
DIGITAL ASSET INTELLIGENCE + EXECUTION
EXECUTED BY
BB DIGITALNA AGENCIJA

Investigation, consulting and execution of digital assets, premium-domain strategies, information architecture, semantic systems, websites and agreed digital growth plans.

TOPICALAUTHORITY.ORG / SEMANTIC INTELLIGENCE SYSTEM BB DIGITALNA AGENCIJA / BB.HR