PRISM: A Structured Measurement Instrument for Multi-Dimensional Brand Perception

Measuring brand perception requires instruments that capture multi-dimensional structure rather than collapsing it to a single equity score. This paper specifies PRISM (Perception Response Instrument for Structured Measurement), a family of standardized instruments for eliciting multi-dimensional brand perception from both artificial intelligence and human observers within the Spectral Brand Theory framework. The paper makes three contributions. First, it formalizes the PRISM architecture as a five-layer scaffold (PL0 specification, PL1 configuration, PL2 prompts, PL3 sessions, PL4 analysis) that is domain-neutral and reproducible. Second, it provides the complete PRISM-B (Brand) specification: eight scale items mapped one-to-one to SBT dimensions, a 1–5 ordinal response format empirically demonstrated as the minimum-distortion operating point for AI observers across 17 large language model architectures (with human validation pending), an exact prompt template, and a scoring algorithm comprising the Dimensional Collapse Index and cross-observer cosine convergence. Third, it situates PRISM-B within a measurement family (PRISM-M for metamerism discrimination, PRISM-T for temporal tracking, PRISM-C for choice prediction) and establishes connection to multi-observer triangulation via Perception DOP. The instrument is specified here; validation against human-subject data is a separate future study. Includes zharnikov-2026-prism-instrument.yaml (Paper Spec v0.1.0) – a machine-readable specification of the paper's claims, assumptions, and dependencies. The paper's full machine-first bundle (the SPINE claim/dependency graph and the ONTOLOGY term module) lives in the public repository; see https://github.com/spectralbranding/paper-spec for the standard. This PDF is generated programmatically from that machine-first source under a research-as-repository model.