Data Standard · BCI Lexicon
Data Reliability Grade
DRG
The Data Reliability Grade (DRG) framework grades every input to a BCI reading by its evidentiary reliability — disclosed alongside the reading it supports, and never silently averaged away.
The DRG Scale
| Grade | Name | Weight | Example Sources |
|---|---|---|---|
DRG-A | Hard Audited Data | 1.0 | 10-K/10-Q filings, government statistics, real-time exchange data |
DRG-B | Structured Observational Data | 0.8 | Bloomberg, Nielsen, SimilarWeb |
DRG-C | Unstructured Semantic Data | 0.5 | Social-media keyword heat, NLP-cleaned sentiment |
DRG-D | Anecdotal Evidence | 0.0 | Expert interviews, internal hearsay — footnote only |
Circuit Breakers
- Data-Missing Circuit Breaker: any dimension lacking DRG-A or DRG-B data is marked N/A. Never filled by estimation or averaging.
- Logic-Conflict Circuit Breaker: an internally contradictory reading (e.g. MT rising while TS falls sharply) triggers mandatory manual review before publication.
One consolidated grade for a reading is never presented as if all its inputs shared the same reliability. Every published BCI reading discloses its DRG grade per variable, individually.
Definitional Authority — This term is defined canonically in BSIP v4.0 and the White Paper v4.0. This page restates that definition on its own address for reference; it is not an independent source and does not diverge from it.
Rating Limitation — Nothing on this page constitutes a credit rating, valuation opinion, or investment instruction of any kind.