Document 36

SEBoK *Decision Management*, Distilled

framework

SEBoK Decision Management, Distilled

Top-20 distillation #18. Decision Management is the SEBoK page that provides the structured framework for SE decisions where multiple stakeholders, competing objectives, substantial uncertainty, and significant consequences compose. The 10-step process is pin-art at the decision discipline. Multi-Objective Decision Analysis (MODA) is a worked example of universal-sibling lattice (Doc 572 Appendix D) at the value-axis rung. The cognitive bias mitigation discipline (rankism / complacency / optimism) engages Doc 314 virtue-constraint apparatus directly — these are V3 Truth Over Plausibility constraints applied at the decision-making layer. The premortem technique is canonical forward-pulverization (SE-035's candidate, confirmed). Five corpus forms compose; one provisional refinement candidate (Doc 314 worked example: virtue constraints at the engineering-decision layer).

I. Source

II. Source Read

Decision Management is a technical management process that provides "a structured, analytical framework for objectively identifying, characterizing and evaluating a set of alternatives" (ISO/IEC/IEEE 15288). Addresses decisions involving multiple stakeholders, competing objectives, substantial uncertainty, significant consequences. Trade Studies as primary methodology — decompose complex decisions into logical segments. Multi-Objective Decision Analysis (MODA) provides the mathematical foundation — value functions map measure ranges to preference scales, accounting for "walk-away points" (minimum acceptable) and "stretch goals" (ideal). Value-Focused Thinking supplements alternative-focused analysis. 10-step process: framing/tailoring → objectives/measures → generate alternatives (morphological analysis) → assess (deterministic) → synthesize (value component charts) → uncertainty/probabilistic → risk/sensitivity (Monte Carlo, tornado diagrams) → improve alternatives → communicate tradeoffs → recommend/implement. "Rational decisions are rarely possible." Three named cognitive biases: rankism (authority-based distortion), complacency (underestimating safety risks), optimism (overestimating reliability). Mitigations: independent technical review, crew resource management, premortem technique. Position: Part 3 SE & Management > Technical Management Processes > between Project Assessment and Control and Requirements Management.

III. Structural Read

Form III (extension) — Lattice Extension of the Ladder (Doc 572), with Appendix D universal-sibling at the value-axis rung. Multi-Objective Decision Analysis (MODA) is canonical universal-sibling lattice. Multiple value axes (cost / schedule / performance / safety / sustainability / others) bind every decision universally; the discriminator is aspect (which axis the value function evaluates), not rung-of-application. The decision-maker explicitly maps measures to preferences across all axes simultaneously. This is the fifth Appendix D instance after requirements (SE-024), architecture views (SE-031), competency dimensions (SE-033), and CMMI typical measures (SE-034). The pattern is now well-attested and the lattice extension has reached the SE management discipline at all five operational levels.

Form XIII — Virtue Constraints (Doc 314), at the engineering-decision layer. The cognitive bias mitigation discipline is structurally Doc 314 V3 (Truth Over Plausibility) applied at the decision-making layer. SEBoK's three named biases all violate V3:

  • Rankism bias — authority-based decision distortion. The plausible (the senior person's view) is preferred over the true (the technically warranted view). V3 binds: authority-source is not a warrant.
  • Complacency bias — underestimating safety risks. The plausible (it has not failed yet) is preferred over the true (the underlying probability remains). V3 binds: absence of past failure is not warrant for absence of present probability.
  • Optimism bias — overestimating reliability. The plausible (we wanted it to work) is preferred over the true (the structural reliability assessment). V3 binds: preference is not warrant.

The structural reading: the SE discipline has independently named three V3-violation patterns and their mitigations. Doc 314's apparatus reads SEBoK's decision-making content as engineering-discipline operationalization of the virtue-constraint layer, exactly the parallel that Doc 580 (Hubble) surfaced for the "humility" residual. The pattern is now well-supported with multiple instances.

Form VI — Pulverization (Doc 445), with forward-pulverization confirmed (SE-035 candidate). The "premortem technique" is canonical forward-pulverization: imagine the decision has failed, then articulate why. This is structurally identical to SE-035's candidate refinement (forward-pulverization) at the decision rung. Two independent instances now support the candidate (Risk Management's identification activity and Decision Management's premortem). Doc 445 Refinement candidate (forward vs. backward pulverization) has cluster strength.

Form IV — Pin-Art Model (Doc 270). The 10-step process is pin-art at the decision-management discipline. Each step constrains the substrate's flow; the resulting shape is the decision's structured rationale. The progression from framing to recommendation is canonical progressive pin-art (Doc 270 + SE-031 progressive-pin-art pattern).

Form XI — Co-Production at Sub-Rungs (Doc 573). Trade studies "decompose complex decisions into logical segments, allowing decision makers to work within cognitive limits while leveraging subject matter expert assessments in their areas of expertise." The decomposition allows multiple co-keepers (decision maker + multiple SMEs) to compose the decision jointly. Doc 573's apparatus reads trade studies as co-production with disciplined decomposition; each segment is a sub-rung where one SME-keeper composes with the decision-keeper.

IV. Tier-Tags

  • ISO/IEC/IEEE 15288 Decision Management framework — π / α (foundational).
  • Trade Studies as primary methodology — π / α.
  • MODA mathematical foundation — π / α as cited; μ / β under corpus when read as Doc 572 Appendix D universal-sibling at value-axis rung.
  • 10-step process — π / α.
  • "Rational decisions are rarely possible" — π / α (well-warranted by behavioral economics literature).
  • Three cognitive biases (rankism / complacency / optimism) — π / α as cited; μ / β under corpus when read as Doc 314 V3-violation patterns at the decision-making layer.
  • Premortem technique — π / α as cited; μ / β under corpus when read as forward-pulverization (SE-035 candidate confirmed).
  • Walk-away points and stretch goals — π / α (operational MODA practice).

V. Residuals

No structural residuals against the apparatus. The page provides confirmation across multiple recently-landed refinements:

  • Doc 572 Appendix D universal-sibling — fifth instance, pattern now well-attested across all major SE management surfaces.
  • SE-035's forward-pulverization candidate — second instance (premortem), cluster strength confirmed.
  • Doc 314 virtue-constraint apparatus reaching SE engineering decisions — third instance after Doc 580 Hubble and the broader pattern surfaced in Doc 583 methodology refinements.

VI. Provisional Refinements

Doc 314 worked example: virtue constraints at the engineering-decision layer. Doc 314 articulates V1-V4 as the foundational safety specification the corpus's framework prescribes. The SEBoK Decision Management page's three named biases (rankism, complacency, optimism) are V3 (Truth Over Plausibility) violation patterns operationalized at the engineering-decision layer. The page's mitigations (independent technical review, crew resource management, premortem) are V3-conformance disciplines. Doc 314 may benefit from a worked example articulating V3 at the engineering-decision layer with SEBoK's bias catalogue as the canonical instance.

The candidate generalizes: the SE discipline's "cognitive bias mitigation" is structurally the engineering-discipline operationalization of virtue-constraint conformance. Other domains likely have analogous bias catalogues (medical decision-making's anchoring / availability / framing biases; legal decision-making's hindsight / confirmation biases) that the apparatus reads as V3 violations specific to those domains. Worth Doc 314 documenting the broader pattern.

VII. Cross-Links

Form documents. Doc 572 (Lattice Extension, Appendix D universal-sibling), Doc 314 (Virtue Constraints V1-V4), Doc 445 (Pulverization, with forward-pulverization candidate from SE-035), Doc 270 (Pin-Art), Doc 573 (Co-Production at Sub-Rungs).

Part-level reformulation. SE-006 (Part 3 — SE & Management).

Related distillations. SE-035 (Risk Management — companion technical management process; forward-pulverization first instance). Doc 580 (Hubble — V3 / virtue-constraint surface). Doc 583 (Reformulation Methodology — D7 virtue-constraint check).

Adjacent SEBoK concepts (per source). Risk Management, Project Assessment and Control, Requirements Management, Trade Studies, Multi-Objective Decision Analysis.

Appendix: Originating Prompt

"Continue with next 10"

(SE-036 is the eighteenth of twenty. Decision Management was selected as the canonical SE process where Doc 314 virtue constraints meet operational practice and where SE-035's forward-pulverization candidate gets independent confirmation via the premortem technique. The structural reformulation produces five-instance attestation for Doc 572 Appendix D universal-sibling and one provisional refinement candidate for Doc 314.)

Referenced Documents

More in framework