SNAP Applicability Scorecard
Evidence Decision Weight Assessment
Validity asks: did the study measure what it claims?
Applicability asks: how much decision weight should I assign this evidence for my context?
New here? Read the full SNAP Applicability Scale for level definitions, thresholds, and a worked example.
Provenance note: higher P is not better. This axis measures distance from unmediated reality, not quality. P4 (experimental) is more constructed than P1 (observed), but often produces stronger attribution. These are independent properties. Select multiple P levels for hybrid data; the (H) flag attaches automatically.
Relative to the domain's rate of change, not calendar age. In AI tooling a generation is 12 to 18 months.
Independently confirmed by another team, context, or population? Affects policy and standards thresholds.
SNAP Signature
Modifiers
—
Computed Evidence Grade
Rate all four axes to compute