DICE Theory: A Decision Integrity and Completeness Evaluation Framework

Decision-making in complex, uncertain, and volatile environments requires not only robust cognitive processing but also structured evaluation of decision quality. This work introduces DICE Theory (Decision Integrity and Completeness Evaluation) as a novel evaluative construct designed to assess decision quality across four dimensions: Depth, Integration, Coherence, and Exhaustiveness.

DICE Theory addresses a critical gap between traditional focuses on decision processes (how decisions are made) and performance metrics (what outcomes are achieved). It evaluates how structurally sound, integrated, and complete a decision is before execution, making it particularly relevant for volatile, uncertain, complex, and ambiguous (VUCA) contexts and human–AI decision environments.

Positioned within a broader decision intelligence stack, DICE acts as an intermediate evaluation layer between cognitive processing models (e.g., EIARA/TIER) and quantitative measurement systems (e.g., DMQS). EIARA and related cognitive frameworks explain how decisions emerge, DICE evaluates the integrity and completeness of the resulting decisions, and DMQS focuses on how those decisions perform in practice.

The four DICE dimensions capture distinct but interdependent aspects of decision robustness. Depth concerns analytical and contextual rigor, including data exploration, scenario analysis, and critical reasoning. Integration reflects the alignment of emotional, intuitive, and analytical inputs into a harmonised decision. Coherence focuses on internal logical consistency, ensuring assumptions, reasoning, and conclusions are structurally defensible. Exhaustiveness, aligned with the MECE principle (mutually exclusive, collectively exhaustive), assesses coverage of relevant variables, risks, constraints, and alternative pathways.

DICE Theory contributes to decision intelligence and explainable AI (XAI) by providing human-interpretable criteria for judging decision quality beyond standard AI metrics such as accuracy, precision, and recall. It supports transparency, trust, and accountability in human–AI collaboration by making the structural and cognitive quality of decisions explicit and assessable.

Potential application domains include strategic decision-making (business strategy, market entry, investment planning), operational decision-making (process optimisation, resource allocation, performance management), and human–AI collaboration (AI-assisted recommendations, decision support systems, hybrid intelligence models). The framework can be operationalised into measurement instruments and integrated into mixed-methods research designs to empirically study decision quality.

Future work includes developing validated measurement scales for each DICE dimension, empirical testing across industries (such as telecom SMEs, healthcare, and finance), integration with AI-driven decision systems, and mapping DICE to established theories such as bounded rationality, dual-process theory, and XAI frameworks. By situating DICE between cognitive processing and quantitative measurement, this work lays the foundation for a full-stack decision intelligence architecture spanning EIARA/TIER → DICE → DMQS.