Early access — cascade metrics are real (derived from canonical token telemetry); the operator field is a curated seed. Learn more about the data

Task Conditioning

0Concept

Whether metrics are conditioned by task type — different tasks may produce systematically different metric profiles regardless of operator skill.

Last updated: 2026-09-01Spec: SigRank Standard v1.0 (proposed)

Definition

Whether metrics are conditioned by task type — different tasks may produce systematically different metric profiles regardless of operator skill.

Inputs

Metric profiles for the same operator across different task types.

Derived variables

Task-level metric differences, task effect size.

Claim

Metrics may be task-dependent — the same operator may show different metrics on different tasks, and this is not a skill difference.

Test

Compare the same operator's metrics across task types. Test for significant task effects.

Observable

Task-level metric differences, task effect F-ratio.

Falsifier

No task effects — metrics are identical across all task types for all operators.

Evidence

Field data suggests task effects exist. Formal analysis not yet conducted.

Limitations

Task classification is subjective. Some tasks may be rare in the dataset.

Lineage

Upsilon workflow friction analysis, proposed validation. Architecture: mos2es.com/architecture.

Cross-references