Skip to content

R0050/2026-03-31/Q002/SRC05/E01
Research R0050 — Journalism and Other Truth-Seeking Disciplines
Run 2026-03-31
Query Q002
Source SRC05
Evidence SRC05-E01
Type Factual

FMEA's three-axis scoring methodology is a novel multi-dimensional risk quantification approach.

URL: https://www.6sigma.us/six-sigma-articles/risk-priority-number-rpn/

Extract

FMEA assesses each failure mode on three independent axes, each scored 1-10:

  1. Severity (S): How serious are the consequences if the failure occurs?
  2. Occurrence (O): How likely is the failure to happen?
  3. Detection (D): How likely is the failure to be detected before impact?

RPN = S x O x D (range: 1 to 1000).

Novel concept: Detection as a distinct evaluation dimension. None of the nine frameworks formally evaluate whether an error, bias, or gap would be detected by the process. ROBIS audits process quality but does not score detection probability. ICD 203 assesses confidence but not detectability of analytical errors. FMEA introduces the concept that a finding's risk depends not just on its likelihood and impact but on whether the analytical process would catch it if it were wrong.

Additional novel concept: Multi-axis quantitative risk scoring. The nine frameworks use qualitative assessment (High/Medium/Low, probability bands) or single-axis quantification (ICD 203 probability scale). FMEA's three-axis approach produces a composite score from independent dimensions.

Caveat: The RPN has known mathematical limitations — ordinal scale multiplication can produce rank reversals. The newer AIAG-VDA FMEA replaces RPN with action priority (AP) tables, suggesting the field itself recognizes this limitation.

Relevance to Hypotheses

Hypothesis Relationship Strength
H1 Supports FMEA contributes genuinely novel concepts
H2 Contradicts Novel concepts exist
H3 Supports FMEA is one of the few disciplines with genuinely novel contributions

Context

FMEA's detection dimension is potentially the most transferable novel concept for research methodology. A research process that evaluates not just "is this finding likely correct?" but "would we detect it if it were wrong?" adds a quality layer not present in the existing nine frameworks. The self-audit (ROBIS) partially addresses this but does not formalize detection probability.