`expected_results`¶

The expected_results section documents the outcomes that should follow a successful execution-qualitative descriptions and quantitative metrics that support experiment verification.

Field Reference¶

Field	Type	Required	Units	Allowed Types	Constraints / Enum	Example
`description`	string	✅	-	string	Outcome summary	`"Supernatant contains lysed DNA."`
`quantitative_metrics[]`	list\	optional	various	array	Each `{ name, value, unit }`	`[{ name: "DNA_yield", value: 20, unit: "µg" }]`
`method`	string	optional	-	string	Instrument or assay used	`"Spectrophotometer"`
`confidence_level`	string	optional	-	enum	`high`, `medium`, `low`, `unknown`	`"high"`

Metric Structure¶

Quantitative metrics should be explicit and machine-readable:

quantitative_metrics:
  - name: "Purity_ratio"
    value: 1.8
    unit: "A260/A280"
    target: ">= 1.8"
    tolerance: "±0.1"

Supply canonical SI units whenever possible; for ratios such as A260/A280, treat the unit as dimensionless but still provide a descriptive string.
Include optional target, tolerance, or notes keys to communicate acceptable variance bands.

Confidence Levels¶

high - empirical result reproduced consistently with strong evidence.
medium - outcome observed but with moderate variance or smaller sample size.
low - exploratory or preliminary data; treat deviations cautiously.
unknown - no reliability statement can be made (e.g., new protocol).

Best Practices¶

Connect method to the device or assay actually used (e.g., "LC-MS" or "Flow cytometer"). When possible, match it to an entry in devices by ID via notes.
Provide supplemental plots, raw data, or analysis notebooks through the attachments section and reference them in expected_results.notes.
Use confidence_level to inform downstream dashboards which metrics can gate quality checks.
Document statistical treatments (mean, median, confidence intervals) either in notes or attached analysis files for reproducibility.

Validation Notes¶

Every metric entry requires a name; validators reject unnamed objects.
Numeric value fields should be accompanied by unit unless dimensionless. Missing units raise Q-class warnings in strict mode.
Enumerated confidence_level values are validated verbatim-other strings are rejected.

expected_results¶