Research
Our numbers need to earn your trust.
Here's where AvgPay's evidence comes from, which rules decide what gets published, and the uncertainty we won't hide.
Data sources and publication scope
| Evidence | Currently loaded scope | What it does not prove |
|---|---|---|
| Occupational wages | BLS OEWS, May 2025; 18 occupations in the validated release | Individual offer, specialty wage, employer grade or total compensation |
| Work and skills | Source-verified O*NET profiles, version 31.0 | Personal competence or observed career transitions |
| Employment outlook | 2025–2035 BLS projections | Current open vacancies, personal hiring chances or a forecast guarantee |
| Regional purchasing power | BEA regional price parities, 2024 | Household spending, taxes, individual net income or a live cost-of-living quote |
Publication and correction safeguards
Our automated pipelines reject invalid source identities, broken joins, missing essential observations, duplicate occupation slugs, unsupported salary-level claims and inconsistent methodology inputs. Static builds and page-level gates prevent browse-only occupations from automatically becoming indexed SEO guides.
Automated checks have limits. They do not substitute for external editorial review, employer validation or professional advice. We do not represent mechanically generated prose as independently expert-reviewed and do not create fake author biographies or qualifications.
If a source record changes or a metric fails validation, our policy is to withhold or qualify the affected insight and preserve the last verified release where appropriate, rather than silently filling values. Corrections should update the affected source or model and its dependents. A newly displayed date must reflect an actual substantive update, not a cosmetic SEO refresh.
How to flag a correction
For a question about a source, calculation or statement, email support@avgpay.com with the page URL, the figure or claim you are questioning, and any relevant published evidence. Do not include sensitive personal salary information. Corrections should be assessed against the source vintage and documented methodology before publication changes.
What counts as AvgPay's original analysis?
Our proprietary value is the documented blending, comparison and interpretation of published evidence: occupation pay-position metrics, skill relationships, role-progression frameworks, regional wage-price adjustments and reproducible research questions. Original analysis does not mean we collected every underlying number, and methodology links retain primary-source attribution.
Research artifacts should explain eligibility rules, observation counts, source releases and limitations. Where a table can be reproduced, we publish the filter or calculation and, where rights permit, a downloadable dataset. See our location-price research for an example.
How to interpret the most consequential estimates
Job-level responsibility ladders describe illustrative routes, not personnel records or mandatory advancement stages. AI task-exposure estimates are not layoff probabilities. National wage percentiles are not entry-, mid- or senior-level wages. Forecasts are modeled scenarios with assumptions and uncertainty, not official wage releases.
When evidence does not support a claim, AvgPay should show “not established” instead of a made-up number.
About AvgPay → · Detailed methodologies → · Current original research →