Primary-source datasets
Research begins with attributable government records, statutes, agency publications, and documented methodology.
Evidence desk
Research designed to show its work.
Every published analysis identifies its source period, calculation method, exclusions, limits, review date, and corrections path.
Reproducible review
Published findings will identify the period examined, definitions used, exclusions, and material limitations.
Independent interpretation
Commercial relationships do not determine findings, rankings, or the selection of evidence.
Published analysis
IRS Offer in Compromise acceptance trends
A source-labeled review of offers received and accepted, with explicit warnings against using aggregate rates to predict an individual result.
Reader journey testing protocol
Moderated sessions begin with a realistic prompt, such as receiving an IRS notice, without naming a relief program. We measure whether a reader can identify the deadline, explain the immediate risk, find a suitable next action, and distinguish education from professional advice.
- Recruitment: Mix of notice recipients, self-employed taxpayers, and readers with no tax vocabulary.
- Success: Correct path and next action without moderator assistance.
- Safety: No collection of SSNs, tax transcripts, exact balances, or account credentials.
- Iteration: Log confusion by page and revise labels before adding features.
Before a study is published
Read the standards that govern source selection, checking, corrections, and any use of automation.