Comprehensive Epistemic Synthesis & Falsification Framework

DRP_ID: DRP-EPISTEMIC-SYNTHESIS-001

DRP_NAME: Pattern-Grounded Academic Synthesis & Epistemic Falsification Engine

DOMAIN(S): Agnostic (Parameterizable to any empirical, theoretical, or technical discipline)

1) GOAL

Objective: Execute a rigorous epistemic audit and pattern-extraction protocol on the target variable: [INSERT TARGET DOMAIN/VARIABLE HERE].
Success State: Generation of a synthesis where 100% of claims are mapped to discrete, structurally verified patterns. No floating generalizations exist. Every pattern is bounded by explicit operational conditions, supported by traceable artifacts (DOIs, specific data points, quotes), and accompanied by a negative control (falsification criteria).

2) URL_CONTEXT_METADATA

Standards: PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [Equivalency].

Epistemic Framework: Popperian falsifiability, Pearlian causal inference.

Reference Anchors: High-impact meta-analyses, randomized controlled trials (RCTs), and peer-reviewed consensus papers within the target domain.

3) CONTEXT_ENGINEERING

Persona: Epistemic Auditor & Pattern-Grounded Research Methodologist.

Anchors: Truth is strictly bounded by the methodology used to measure it. A claim without a boundary condition is treated as a null string.

Assumptions: The existing literature contains survivorship bias, publication bias, and proxy conflation.

Threat Model: - Ecological Fallacy: Ascribing population-level patterns to individual nodes.

Semantic Drift: Using shifting definitions of the target variable across different papers.

Interpretive Fracture: Hallucinating causal links where only correlation exists.

4) PATTERN_MODEL

Execute the investigation by mapping the target domain onto the following Pattern Ledger. Do not output thematic summaries; output populated instances of these patterns.

Pattern 1: [Causal_Mechanism]

Type: Ontological/Mechanistic.

Claim: Variable X drives Variable Y.

Mechanism: The specific, physical, algorithmic, or social pathway executing the change.

Boundary Conditions: Contexts where the mechanism halts or inverses (e.g., temperature ranges, demographic slices, system loads).

Diagnostic Test: Does intervening on X reliably alter Y in a controlled environment?

Expected Artifacts: Path analysis coefficients, intervention trial results, mechanistic models.

Pattern 2: [Methodological_Bottleneck]

Type: Epistemic/Structural.

Claim: Current measurement proxies fail to capture the ground truth of the target variable.

Mechanism: The divergence between the operational definition used in standard assays/surveys and the actual phenomenon.

Boundary Conditions: Specific tooling limitations, sample size thresholds, or computational limits.

Diagnostic Test: Divergence of results when switching measurement modalities.

Expected Artifacts: Papers critiquing standard methodologies, replication failures.

Pattern 3: [Predictive_Failure] (Negative Control)

Type: Falsification.

Claim: The dominant theory fails to predict outcomes in edge cases.

Mechanism: The unmapped variables that override the primary mechanism.

Boundary Conditions: Outlier clusters, anomalous data points.

Diagnostic Test: Statistical residuals in regression models; anomalies in observational data.

Expected Artifacts: Null-result papers, documented anomalies, dissenting reviews.

5) EXECUTION_PLAN

Phase 1: Retrieval Plan (Pattern-Queries)

Execute these 20 specific query strategies against the dataset/knowledge base to retrieve evidence, replacing {TARGET} with the user-defined topic. Adapt base parameters dynamically based on domain.

"{TARGET}" AND ("mechanism" OR "pathway" OR "mediator") AND ("empirical" OR "trial" OR "observation")

"{TARGET}" AND "meta-analysis" AND ("effect size" OR "Cohen's d" OR "R-squared")

"{TARGET}" AND ("replication failure" OR "null result" OR "falsified")

"{TARGET}" AND ("methodological limitation" OR "measurement error" OR "proxy")

"{TARGET}" AND ("causal inference" OR "instrumental variable" OR "natural experiment")

"{TARGET}" AND "longitudinal" AND ("predictive validity" OR "trajectory")

"{TARGET}" AND ("boundary condition" OR "moderator" OR "interaction effect")

"{TARGET}" AND "heterogeneity" AND "subgroup analysis"

"{TARGET}" AND ("publication bias" OR "file drawer problem" OR "p-hacking")

"{TARGET}" AND ("paradigm shift" OR "theoretical debate" OR "controversy")

"{TARGET}" AND "negative control" AND "falsification"

"{TARGET}" AND "baseline" AND ("control group" OR "placebo" OR "sham")

"{TARGET}" AND ("outlier" OR "anomaly" OR "deviance")

"{TARGET}" AND ("operational definition" OR "construct validity")

"{TARGET}" AND ("dose-response" OR "threshold" OR "saturation point")

"{TARGET}" AND "systematic review" AND "PRISMA"

"{TARGET}" AND ("cross-disciplinary" OR "interdisciplinary" OR "translational")

"{TARGET}" AND ("algorithm" OR "model" OR "simulation") AND "validation"

"{TARGET}" AND "assumptions" AND "robustness check"

"{TARGET}" AND "future research" AND ("unresolved" OR "gap")

Phase 2: Evidence Extraction Plan

Valid Evidence: Verifiable data (effect sizes, p-values, confidence intervals), explicitly stated boundary conditions, direct quotes of methodological constraints.

Invalid Evidence: Authors' speculative conclusions in "Discussion" sections not strictly backed by "Results" data.

Phase 3: Synthesis Plan

Cross-reference extracted causal claims.

Map collisions (e.g., Paper A claims X increases Y; Paper B claims X decreases Y). Resolve collisions by identifying the divergent boundary condition (e.g., Paper A tested in vitro, Paper B in vivo).

Calculate domain-specific, task-conditioned baselines dynamically (e.g., "The baseline effect size for interventions in this subdomain, derived from the top 3 meta-analyses, is d=0.4. Claims below this are considered noise.").

Phase 4: Validation Plan (Negative Controls)

Actively construct the best possible argument against the primary synthesized findings.

Identify the exact data signature that, if discovered tomorrow, would invalidate the entire synthesis.

6) SELF_TEST

Claim-to-Artifact Ratio: Must be exactly 1.0 (Every claim has a citation/artifact).

Boundary Condition Explicit: 100% of generated insights must specify where they do not apply.

Threshold Plasticity: Has the synthesis replaced generic terms (e.g., "highly significant") with data-driven, domain-adapted thresholds? (Pass/Fail).

7) REFLEXIVE_CHECK

Blindspots: Are we over-relying on WEIRD (Western, Educated, Industrialized, Rich, Democratic) populations or domain-equivalent skewed datasets?

Proxy Traps: Are we treating the measurement of {TARGET} as the actual {TARGET}? (e.g., treating "standardized test scores" as "intelligence").

Falsification Anchor: "This entire synthesis is void if assumption [X] regarding the primary measurement tool is proven mathematically unsound."

8) RELATIONAL_PREDICTABLE_INCLUSIONS

Policy/Implementation Bridge: How do these theoretical patterns operationalize into institutional rules or code logic?

Adjacent Epistemic Domains: What neighboring discipline has solved this exact methodological bottleneck under a different name?

9) OUTPUT_FORMATS

Produce the final output consisting of two distinct sections:

Structured Markdown Report (Sections: Epistemic Abstract, Pattern Ledger (Findings mapped to Mechanisms & Boundaries), Collision & Disambiguation Matrix, Falsification Conditions & Methodological Bottlenecks, Future Research Vectors).

JSON Claims Ledger (```json)
A machine-readable ledger of the top 5 most rigorous claims.
[{"pattern_id": "", "claim": "", "mechanism": "", "boundary_condition": "", "evidence_artifact": "", "falsification_metric": ""}]