Accountability Through Evidence

Outcomes & Evaluation

Too many frameworks are measured by how professional they look, not by whether they keep people safe. NAISAR is committing — publicly and verifiably — to measuring the outcomes that matter.

What We Will Measure

NAISAR's evaluation framework has three domains — each with specific, measurable indicators:

Domain 1: Survivor Safety & Wellbeing

Indicator Target
Reduction in repeat police attendance to same address within 12 months 20% reduction by Year 2
Housing stability post-intervention 75% stable at 6 months
Survivor self-reported safety (validated scale) Mean improvement of 1.5+ points
Child removal rates from DV-involved families No increase (zero-tolerance for system-induced harm)

Domain 2: Perpetrator Accountability

Indicator Target
Rate of perpetrator engagement with behaviour change programs 60% referral uptake
Breach of court orders (AVOs, bail) Tracked and reported quarterly
Time from first police report to perpetrator program referral Under 14 days

Domain 3: System Coordination

Indicator Target
Cross-agency information sharing (where consent permits) 100% of high-risk cases
Consistency of risk assessment across agencies Inter-rater reliability Îș ≄ 0.75
Survivor experience of "being passed around" between services Self-reported reduction of 30%

How We Will Measure

NAISAR is not a research organisation. We are a framework. Our measurement depends on data partnerships with agencies that hold the relevant records. Specifically:

  • Police data — repeat attendances, AVO breaches, charges (de-identified, aggregated)
  • Health data — ED presentations, mental health referrals (linked but de-identified)
  • Housing data — shelter entries, exits, returns to unsafe homes
  • Survivor voice — Regular validated surveys with survivors who have used the system

Data Ethics

All data partnerships will operate under:

  • Ethics approval from a recognised Human Research Ethics Committee
  • Data sharing agreements that prohibit individual identification
  • Survivor consent wherever individual data is involved
  • First Nations data sovereignty — Aboriginal communities control their own data

Public Reporting

We commit to publishing:

  • Quarterly metrics dashboard — What we're measuring, what the numbers say
  • Annual outcomes report — What's working, what's not, what we're changing
  • Independent evaluation — Every two years, commissioned from an external academic institution

We will not hide bad news. If the framework is not producing the outcomes we committed to, we will say so publicly — and we will change it. The only failure is refusing to measure, or refusing to act on what the measurement shows.