A restorative justice program should be evaluated through a balanced view of delivery, repair, participant experience, safety, equity, and longer-term patterns—not incident counts alone.
Start with a small, consistent scorecard, then add survey platforms, case-management tools, or independent evaluation support only when reporting needs and case sensitivity justify the effort.
This approach helps leaders see whether facilitation, staff training, and follow-up processes are being delivered as intended. It also prevents a high number of completed conferences from being mistaken for meaningful repair.
The right evaluation method depends on caseload, confidentiality risks, internal capacity, and the decisions the organization must make. Paid tools or outside support can be useful, but only when they improve secure workflows, feedback quality, or accountability.
At a Glance
- Measure repair, safety, fairness, and implementation quality, not only referrals or repeat incidents.
- Separate outputs, such as conferences delivered, from outcomes, such as completed agreements and feeling heard.
- Match the evaluation method to caseload, confidentiality, reporting obligations, staff capacity, and budget.
| Evaluation option | Best fit | Main strength | Key caution |
|---|---|---|---|
| Internal spreadsheet tracking | Small or early-stage programs | Simple tracking of participation, sessions, agreements, and follow-up | Needs consistent definitions and secure access controls |
| Survey or case-management platform | Recurring reporting and larger caseloads | Structured follow-up, surveys, and data-management workflows | Check privacy settings, permissions, and fit with sensitive records |
| External evaluator or data consultant | High-stakes, funded, or public-facing programs | Independent analysis and clearer reporting methods | Define scope, data access, confidentiality, and useful deliverables first |
What Successful Restorative Practice Evaluation Should Show
The Short Answer: Measure Repair, Safety, Fairness, and Implementation—Not Just Incident Counts
A useful evaluation asks more than, “Did referrals go down?” Restorative practices may include circles, facilitated dialogue, conferencing, harm-repair agreements, and follow-up support. A single number cannot fully show whether people felt safe, whether participation was voluntary, whether an agreement was realistic, or whether the process was delivered fairly.
Use a scorecard that combines activity counts, participant experience, repair progress, safety, and equity review. This gives program leaders a more honest basis for decisions about facilitator training, staffing, and program design.
Start With the Purpose of the Program and the Decisions the Data Must Support
Before choosing metrics, write down the decisions the evaluation must support. A school may need to review discipline processes and family communication. A workplace may want to understand whether conflict-resolution pathways are being used. A community program may need to monitor repair plans and follow-up access.
Keep the question practical: What would we change if this measure moved in the wrong direction? If no action would follow, the metric may create reporting work without improving the program.
Use a Small Baseline Set of Measures Before Adding Complex Reporting
Begin with measures your team can define and record consistently. For example, track referrals received, participation, completed sessions, agreement completion, and follow-up status. Record the same definitions over time and use comparison periods carefully.
Do not build a large dashboard before staff can reliably capture the basics. Inconsistent records can create misleading comparisons, even when the reporting software looks polished.
Build a Balanced Scorecard for Restorative Outcomes
Delivery Measures: Referrals, Participation, Timeliness, and Completed Sessions
Delivery measures show whether the program is functioning. Useful examples include referrals, attendance, participation rates, time to resolution, and the number of circles or conferences completed. These are outputs: they describe what was delivered, not whether repair occurred.
Repair Measures: Agreement Completion, Follow-Up, and Participant-Defined Progress
Where harm-repair agreements are used, record whether follow-up occurred and whether agreed actions were completed. Voluntary feedback can also reveal what participants consider progress. One person may value an apology, another may value a practical repair step, and another may simply value being heard without pressure.
A completed agreement should not be treated as proof that all harm has ended. It is one piece of evidence that should be read alongside safety and experience measures.
Experience Measures: Voice, Respect, Voluntariness, and Perceived Fairness
Short, voluntary surveys or open-ended questions can assess whether participants felt heard, treated respectfully, and able to make choices. Facilitator observations and documented themes from debriefs can add context, especially when a number alone hides important concerns.
Use care with power dynamics. Feedback collection should not make a student, victim, accused person, employee, or community member feel pressured to provide a positive response.
Safety and Equity Measures: Access, Exclusion Patterns, and Unintended Harm
Review who is referred, who participates, who declines, who completes agreements, and who receives other disciplinary consequences. An equity review can identify differences in access, participation, outcomes, or exclusion patterns across groups.
Safety questions matter throughout the process. If feedback or records could expose sensitive conflict details, redesign the collection method before expanding it.
Compare Evaluation Options by Effort, Privacy, and Budget
Simple Internal Tracking for Small or Early-Stage Programs
A protected spreadsheet and a small set of standardized forms may be enough when caseloads are manageable and the team needs routine operational information. This option can support regular reviews of referrals, participation, completed sessions, agreements, and follow-up.
It works best when one person owns definitions, staff receive basic documentation guidance, and identifiable records are not broadly shared.
Survey and Case-Management Tools for Recurring Program Reporting
Survey platforms and case-management tools may be worth considering when recurring follow-up, permission controls, or reporting volume increase. They can help standardize questions and reduce manual reporting work. However, a tool does not solve unclear measures or weak facilitation practice.
Before selecting evaluation software, examine data access, retention settings, export controls, user permissions, and secure handling of sensitive information.
When an External Evaluator or Trained Data Consultant Adds Value
External evaluation support can add value when funders, public accountability, or high-stakes decisions require a more independent view. A trained evaluator may also help a program develop consistent definitions, review comparison periods, analyze qualitative themes, and avoid overclaiming impact.
External support is not automatically necessary. It should be proportionate to the program’s sensitivity, reporting obligations, internal skills, and available budget.
What Level of Evaluation Do You Need?
- Is the caseload small enough for secure internal tracking?
- Do you need recurring reports, structured follow-up, or multiple staff permissions?
- Are cases sensitive enough to require stronger data-management controls?
- Do funders, boards, or public partners need independent evaluation?
- Would facilitator training improve documentation quality before buying new software?
Collect Evidence Without Compromising Trust or Confidentiality
Define Terms Consistently and Document the Baseline
Define terms such as “participation,” “completion,” “follow-up,” and “repeat incident” before comparing periods. Document the starting point and avoid changing definitions midway through a reporting cycle without noting the change.
Use Voluntary, Trauma-Aware Participant Feedback Methods
Feedback should be voluntary and designed to reduce pressure. Offer concise questions, explain how responses will be used, and avoid collecting details that are not necessary for improvement. Open-ended responses may provide useful themes, but they require careful handling.
Protect Sensitive Records and Limit Access to Identifiable Data
Restorative records can involve students, victims, accused individuals, employees, and highly personal conflict information. Use secure data handling, limit access to identifiable information, and confirm applicable education-record, employment, safeguarding, and data-protection requirements.
Avoid Treating Satisfaction Scores as Proof of Long-Term Impact
A positive experience score can be meaningful, but it does not prove long-term changes in relationships, safety, attendance, referrals, or repeat harm. Report what the evidence supports and identify what remains unknown.
Adapt the Framework to Schools, Workplaces, and Community Settings
Schools: Discipline, Belonging, Attendance, and Family Communication
Schools may track referrals, participation, attendance at restorative sessions, agreement follow-up, and family communication. Review whether access and outcomes differ across student groups. Keep records aligned with local student privacy and safeguarding requirements.
Workplaces: Conflict Resolution, Retention Risk, and Psychological Safety
Workplace programs can monitor whether people use the process, whether sessions are completed, and whether participants report voice, respect, and fairness. Do not assume that feedback results predict retention or psychological safety without appropriate evidence.
Community and Justice Programs: Repair Plans, Follow-Up, and Service Access
Community and justice programs may focus on participation, repair-plan completion, follow-up contact, and access to support services. Because cases can be particularly sensitive, confidentiality and informed consent should shape every measurement decision.
Selection Criteria and Comparison Summary
Choose internal tracking when decisions are simple, caseloads are manageable, and staff can apply shared definitions. Consider a survey platform or case-management tool when secure workflows, follow-up demands, reporting volume, or permission controls increase. Use independent evaluation support when funding, public accountability, or major program decisions require stronger evidence. Before spending on training, facilitation, data-management tools, or evaluation services, check the provider’s official information for privacy controls, scope, and implementation conditions.
Conclusion
Good restorative justice evaluation is not about finding one perfect metric. It is about collecting enough trustworthy evidence to improve practice without compromising safety or confidentiality. A small, balanced scorecard is often the right starting point. Expand the system only when it helps staff make clearer, more responsible decisions.
Useful Information to Keep in Mind
1. Count activity, but do not confuse activity with outcomes. 2. Use voluntary feedback alongside operational data. 3. Review equity as well as overall totals. 4. Keep definitions stable across reporting periods. 5. Protect identifiable information from unnecessary access.
Important Considerations
No evaluation framework can guarantee reduced repeat harm, fewer referrals, lower absenteeism, or lower costs. Suitable benchmarks, reporting periods, sample sizes, and success thresholds vary by setting. Confirm local legal, safeguarding, education-record, employment, and data-protection requirements before collecting or sharing participant information.
Frequently Asked Questions
Q1. What are the most useful indicators for evaluating a restorative justice program?
A1. Use a balanced set: participation, completed sessions, agreement completion, follow-up, participant voice and fairness, safety concerns, equity patterns, referrals, attendance where relevant, and repeat-incident tracking where definitions are consistent. No single indicator is enough.
Q2. Should a school pay for evaluation software or begin with a spreadsheet and anonymous surveys?
A2. A small program may begin with secure internal tracking and voluntary feedback if staff can manage records consistently. Consider dedicated software when reporting volume, follow-up, access permissions, or secure workflow needs become harder to manage manually.
Q3. How can programs collect restorative practice feedback without putting participant privacy at risk?
A3. Make feedback voluntary, collect only information needed for improvement, explain how responses will be used, limit access to identifiable records, and use secure handling procedures. Review local confidentiality, safeguarding, and data-protection obligations before implementation.
