Enum NegativeResultKind 

pub enum NegativeResultKind {
    RegisteredTrial {
        endpoint: String,
        intervention: String,
        comparator: String,
        population: String,
        n_enrolled: u32,
        power: f64,
        effect_size_ci: (f64, f64),
        effect_size_threshold: Option<f64>,
        registry_id: Option<String>,
    },
    Exploratory {
        reagent: String,
        observation: String,
        attempts: u32,
    },
}

v0.49: NegativeResult as a first-class kernel object.

The essay-driven primitive: when an experiment or trial does not support its hypothesis, the substrate has to be able to record what was tried, in what context, with what observed outcome — without silently flipping the corresponding finding’s confidence. Two shapes carry the depositor’s intent:

• RegisteredTrial: pre-registered trial reads out negative on its primary endpoint. Carries power and effect_size_ci so a downstream reader can tell an informative null (CI tightly bounded around zero, adequate power) from an uninformative one (wide CI, low power) — the distinction the essay calls out: “an underpowered null does not poison downstream confidence.”
• Exploratory: wet-lab dead end. Most failures here cannot be statistically bounded; the substrate’s first job is capturing the (reagent, condition, observed outcome) tuple so the next chemist designing a similar synthesis sees the dead end before she runs the experiment.

vnr_<id> is content-addressed over the canonical preimage of the kind-specific fields plus deposited_by and created. NegativeResults link to the findings they bear against via target_findings; review and retraction follow the same proposal -> canonical event -> reducer pipeline as findings.

Variants

RegisteredTrial

Pre-registered trial whose primary endpoint read out negative.

Fields

endpoint: String

Pre-specified primary endpoint (e.g., “CDR-SB change at 18 months”).

intervention: String

Intervention arm description (drug + dose, device + protocol, etc.).

comparator: String

Comparator arm description (placebo, active control, standard of care).

population: String

Population scope: indication, stage, age range, biomarker eligibility.

n_enrolled: u32

Number of participants enrolled (any arm).

power: f64

Statistical power for the primary endpoint, on [0, 1]. Below 0.8 is the “underpowered null” the essay warns about; this field is what lets a downstream reader distinguish that case from an adequately-powered null.

effect_size_ci: (f64, f64)

Confidence interval for the observed primary effect size, (lower, upper) in the trial’s reported units. A CI tightly bracketing zero with adequate power is an informative null; a wide CI under low power is not.

effect_size_threshold: Option<f64>

Pre-registered minimum effect size of interest, in the same units as effect_size_ci. When the CI excludes this threshold, the null is a positive epistemic claim about the absence of clinically meaningful effect. None when no pre-registered MCID was declared.

registry_id: Option<String>

Trial registry id (e.g., “NCT04532333”). Strongly preferred; the registry id is the load-bearing audit trail.

Exploratory

Exploratory wet-lab failure. Captures the (reagent, condition, observed outcome) tuple even when no statistical bound applies.

Fields

reagent: String

Reagent, compound, vector, or perturbation tried.

observation: String

Free-text observed outcome (e.g., “no measurable expression”, “yields plateau at 6%”, “cytotoxicity at all tested doses”).

attempts: u32

Number of independent attempts whose outcome agreed. Single attempts are fine but should be honest about it.

Implementations

impl NegativeResultKind

pub fn canonical(&self) -> String

Compact canonical representation used in the content-address preimage. Stable across protocol versions — adding a new kind must not change existing kinds’ canonical strings.

Trait Implementations

impl Clone for NegativeResultKind

fn clone(&self) -> NegativeResultKind

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for NegativeResultKind

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for NegativeResultKind

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for NegativeResultKind

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.