Validation and Reproducibility Overview
A summary of the validation framework applied to the Naialu Motion Calculus, including the pilot validation record, reproducibility methodology, and the falsifiability standard governing the framework's outputs.
Two-tier validation.
The Naialu Motion Calculus operates under a tiered validation framework that distinguishes between structural validation, which applies to the deterministic computation engine itself, and application validation, which applies to specific use cases tested against external data.
The distinction matters because the two tiers carry different evidentiary weight. Structural validation establishes that the engine produces the same outputs from the same inputs across operators and time. Application validation establishes that specific outputs correlate with measurable external phenomena. The framework is candid about which tier any given claim rests on.
Structural Validation
Applies to the deterministic cipher and engine. Establishes that any qualified operator running the same procedure on the same input produces the same output. Structural validation is the framework's baseline claim. Every output the engine produces inherits structural validation by construction.
Application Validation
Applies to specific applications of engine outputs tested against external data. Establishes that particular structural reads correlate with particular measurable phenomena. Application validation is study-specific. Each application either has a validation record or does not; the framework does not generalize from one to another.
Structural validation of the engine.
The Naialu Motion Calculus converts a full birth name and birthdate into a particle sequence according to a fixed alphabetic cipher. From the particle sequence the engine derives a precise set of structural values, including field state designation, coherence at source, coherence at operation, render rate, propulsion-retention-dissipation decomposition, field saturation, amplitude, velocity, and the directional complexity profile.
The cipher is deterministic. The engine is deterministic. The structural read is bounded by the engine values produced from the input.
Any qualified third-party operator running the same cipher and engine on the same input produces the same field state designations, the same coherence values, the same render bands, and the same propulsion-retention-dissipation decomposition reported in any Institute output.
The structural-validation claim does not extend beyond engine output. It establishes that the engine is reproducible. It does not assert that engine output predicts any specific behavior, outcome, or external phenomenon. Such assertions require application validation against external data, which is the subject of the next section.
The cipher, particle derivation procedure, and the formulas that produce each structural value from the particle stream are protected intellectual property and do not appear in external-facing documents. The reproducibility claim is verifiable under NDA: qualified reviewers can be given access to the computation procedure and the worked derivation for any reported case.
Application validation record.
Specific applications of engine outputs have been tested against external data. Each study is reported with sample size, effect or accuracy, and honest-scope statement where the result is mixed or bounded.
Render Dynamics Pilot
Pilot study testing engine-derived render rate against external behavioral indicators across nineteen subjects. Reported correlation of 0.80 establishes a strong relationship between computed render values and observed behavioral pace within the pilot sample. Sample size is small; the finding supports continued research at larger scale rather than constituting a fully powered validation.
NCAA Tournament Application
Application of institutional architecture analysis to NCAA basketball tournament outcomes. Reported accuracy ranges from seventy-five percent to eighty-three percent depending on bracket round and the specific architectural read applied. The application targets institutional configuration rather than game-level variance; accuracy is highest where institutional architecture is the dominant outcome driver and falls as game-level factors increase in weight.
NBA Game-Level Predictions
Application of structural analysis to NBA game-level outcomes. Reported accuracy is fifty-two percent, barely above chance. This figure is preserved in the validation record as an honest-scope credibility signal. The framework does not claim predictive accuracy at this level of variance, and the result is included specifically to demonstrate that the framework reports both ranges where it works and ranges where it does not.
Where the framework correlates strongly (institutional and architectural), it correlates strongly. Where it does not (single-game variance with high stochastic content), it does not. Reporting both is the validation posture; concealing the lower-accuracy result would compromise the credibility of the stronger results.
Linguistic Corpus Analysis
Structural analysis applied to a documented corpus of five hundred twenty-one thousand nine hundred thirty-six words of subject-produced text. Engine-derived structural signature compared against linguistic patterns measurable in the corpus. Statistical significance at p less than 0.001 establishes that the relationship between computed structural read and observed linguistic patterns is not attributable to chance within the corpus.
What is not yet application-validated.
The framework produces structural reads across many applications that have not been independently tested against external data. Those applications inherit structural validation by construction, since they derive from the same deterministic engine. They do not carry application validation, and the Institute does not represent them as such.
This includes the bulk of single-subject analytical work produced under engagement. A leadership readiness diagnostic, a coherence report, an integrated architecture schematic, a mission mapping read: each is structurally valid by construction (the engine produces the same output from the same input) but is not separately validated against external behavioral data for the specific subject. The structural read is what the engine reports. Whether that read predicts a specific outcome in a specific case is a separate question that requires separate validation.
The framework's honest posture on this distinction is part of its validation discipline. Specific predictive claims are bounded; structural-measurement claims are reproducible. Procurement reviewers evaluating the framework should weight each claim accordingly.
Reproducibility methodology.
Third-party verification of the deterministic claim is supported under controlled conditions. The Institute does not publish the computation procedure but does not prevent verification.
Pre-Registration of Procedure
The cipher and engine procedure are fixed in advance of any specific computation. They are not tuned to produce a particular result on a particular case. The fixed-procedure claim is verifiable: identical inputs produce identical outputs, and operators can confirm this on inputs of their choosing.
NDA-Supported Access
For qualified procurement reviewers, methodology evaluators, and academic research partners, the Institute supports verification access under non-disclosure agreement. Access includes the cipher procedure, the engine formulas, and the worked derivation for any case reported in the Institute's published materials.
IP Boundary on Public Documents
No cipher, particle chain, operator chain, engine formula, or raw computation data appears in any external-facing document, including this overview. The architecture is described at the property level only; the computation procedure remains protected. This boundary is consistent across all Institute documentation and is the basis for the framework's commercial viability.
Auditability for Federal Procurement
For federal procurement contexts requiring full auditability, the Institute supports audit-ready engagement structures including documented computation traces, audit logs of engine versions used in specific outputs, and reviewer access protocols compatible with FAR and DFARS audit requirements.
Falsifiability standard.
The structural-validation tier of the framework is falsifiable under a specific condition: if a qualified third-party operator runs the cipher and engine on the same input and produces a different output, the determinism claim fails. The Institute supports this test under NDA and considers it the primary falsification standard for the engine itself.
The application-validation tier is falsifiable per study. Each application-validated study has reported sample size, effect size, and statistical significance. Replication of any study at equivalent or larger sample size with materially different results would constitute disconfirmation of the specific application claim.
The framework does not protect itself from disconfirmation through ambiguity. Specific predictive claims are bounded with specific numbers. Specific structural-measurement claims are reproducible by specific procedure. Where claims are vaguer (as in some interpretive elements of single-subject reports), the framework is candid that those elements are interpretive rather than computed, and they carry the evidentiary weight of expert interpretation rather than direct measurement.
Engine determinism fails if qualified operators produce divergent outputs from identical inputs. Application validation fails if replication at equivalent or larger sample produces materially different results. Both conditions are testable under controlled access.