Methodology

How Abundera Research measures things, sources data, and decides what to publish.

Pre-registration

Every report is pre-registered before measurement begins. The pre-registration documents the research question, the data sources, the measurement approach, the inclusion and exclusion criteria, and the target release date. The pre-registration is published on this site and dated.

When the aggregate data is collected, the full report ships and references the pre-registration. If the pre-registered methodology turns out to be wrong or impractical, the revision is documented in the published report.

Data sources

Three classes of source feed the research.

• Public datasets. CMS, BLS, KFF, Census, state-level transparency portals. Every figure is sourced to a functional URL.
• Aggregate findings from Abundera tools. Anonymized, deduplicated, and aggregated such that no individual user is identifiable. Sample-size caveats and selection-bias disclosures published with every aggregate figure.
• Reproducible analyses of public records. When a public dataset exists but no one has aggregated it usefully, that is itself a research output worth publishing.

Editorial line

The editorial line is separate from the product. If the data shows the product is not helping in some category, the data ships unchanged. Marketing language stays out of the report body.

Corrections

Substantive corrections are noted in a corrections section on the report itself, dated. Minor copy edits are made silently.

Falsifiability

Every claim in a report is intended to be falsifiable. If a published figure turns out to be wrong, the correction is welcomed: research@abundera.ai.