DFAS-RLG-01: Impact-Based Evaluation and Citation Metrics: An Ethical and Epistemic Invalidity Analysis

This paper introduces a normative governance declaration within Dynamic Financial Applied Meta-Science (DFAS), establishing the ethical and epistemic invalidity of impact-based evaluation systems, including citation metrics, impact factors, and derivative indicators.

Contemporary research evaluation relies extensively on numerical metrics as proxies for scientific quality, legitimacy, and contribution. This study demonstrates that such metrics are not merely biased or misapplied, but structurally ineligible to function as evaluative instruments. Impact-based systems measure diffusion dynamics—visibility, network amplification, and temporal accumulation—rather than epistemic rigor, methodological responsibility, or accountable decision-making.

The paper advances a decisive conceptual separation between three non-substitutable domains:

• Diffusion (knowledge spread)
• Validity (epistemic soundness)
• Responsibility (governed knowledge formation)

By collapsing these domains into a single numerical dimension, metric-based systems conflate visibility with legitimacy, producing ethically indefensible and epistemically invalid judgments.

The study establishes that:

• Impact and citation metrics are non-normative diffusion signals, not indicators of scientific legitimacy
• No reform, normalization, or alternative metric can correct this failure, as the defect is structural rather than technical
• Metric-based evaluation systems displace responsibility, violate proportionality, and produce systemic harm in research governance
• Output-indexed evaluation is incompatible with path-governed legitimacy, which requires traceable, temporally coherent decision processes

Rather than proposing improved metrics, this paper withdraws the evaluative authority of metrics altogether, redefining research legitimacy as a function of governed research paths, temporal traceability, and accountable causal reasoning.

The work further demonstrates how metric-driven systems distort editorial behavior, suppress high-risk research, erase the temporal formation of knowledge, and mis-train artificial intelligence systems by embedding visibility as a proxy for legitimacy.

This manuscript constitutes a governance declaration, not a reform proposal. Its contribution lies in removing an illegitimate evaluative authority and establishing the conceptual conditions for future research governance systems grounded in responsibility rather than numerical aftermath.