# EQIQs™ Bias & Fairness Policy

Last reviewed: 2026-10-02 · Owner: Jessica Michelle Baker · Review cadence: 90 days

## Position
EQIQs™ instruments are decision-support tools, not decision-makers. We
prohibit hiring, screening, ranking, firing, pay and promotion uses, including as one input among others. Recruiter use is post-offer onboarding only.

No EQIQs score is a pass/fail gate. EQIQs does not publish, calculate, or
support cut scores, minimum thresholds, qualifying bands, or selection
rankings, and the product exposes no feature that auto-rejects, auto-advances,
or auto-ranks candidates. Building an external cut score on top of the API,
MCP server, or exports violates the hiring terms.

Full policy: https://eqiqs.com/trust/adverse-impact

## Research basis
- Items are modeled on established, peer-reviewed
  frameworks (Big Five, DISC, MBTI lineage, RIASEC, Enneagram, Attachment
  Theory, Schein Career Anchors). EQIQs itself has not been independently validated as a selection instrument.
- No population norm or standard-error claim is made without published supporting evidence.
- Compatibility scores use role-adjusted weighting and require a minimum
  of 3 completed frameworks per profile (Data Sufficiency Rule) before
  surfacing.

## Protected Traits
- AI prompts (Roast suite, Coaching, Reports) are explicitly instructed
  to never reference: race, religion, age, gender identity, sexual
  orientation, disability, national origin, pregnancy, or political
  affiliation.
- All Roast outputs pass through a profanity + protected-class filter
  before display.

## Disparate Impact
- Aggregate score distributions are reviewed quarterly across self-reported
  demographic segments where users opt-in.
- Any framework showing >0.2 standardized mean difference between groups
  triggers an item-level fairness review.

## Adverse Action
EQIQs does not support selection workflows. Use voluntary results for communication, coaching and post-offer onboarding; consult qualified legal counsel for employment processes. No selection-validity documentation or hiring approval is promised.

## Audit & Transparency
- Customers can request a per-decision explanation report showing which
  frameworks and dimensions contributed to a compatibility score.
- Scoring formulas are versioned. AI prose can vary and exact historical regeneration is not guaranteed.

## Contact
- Fairness concerns: jessica@eqiqs.com
- Methodology questions: BD@EQIQs.com