What is an equity dashboard?
An equity dashboard brings together measures of access, experience and outcomes so a team can examine differences across relevant groups. It should show the population, period, definitions and limitations behind those comparisons—not just a set of colored bars.
A useful view helps reviewers decide what to investigate and what action to consider. It does not determine the cause of a disparity, prove fairness or require public access to individual records.
This guide includes an illustrative layout, a worked group comparison and a practical build process. Start with equity metrics if the measures are not yet defined. For workplace representation and employee listening, use the separate DEI dashboard guide.
Build the view around a review question
For example: “Can eligible participants access and complete the program, and where should we investigate barriers?” This is more useful than asking the dashboard to summarize everything the organization knows.
Scroll horizontally to see all columns →
| Dashboard area | Show | Purpose |
|---|---|---|
| Scope | Program, population, observation period and last review | Make clear what the reader is looking at |
| Coverage | Eligible records, known outcomes and usable group information | Expose the limits before interpreting gaps |
| Group comparison | Counts, rates, reference group and relevant uncertainty | Show differences without hiding denominators |
| Context | Relevant site, program or collection differences | Avoid treating unlike populations as equivalent |
| Evidence review | Definitions, approved summaries and authorized source access | Let reviewers inspect the basis of a finding |
| Action | Question to investigate, owner and next review | Connect the view to a practical decision |
Keep the first screen selective. Additional detail can be available through clearly labelled expansion or a separate review page. The reader should not need to interpret a dozen filters before understanding the main question.
Example dashboard: completion across two groups
Fictional teaching example. The same program and completion definition apply to 150 eligible participants during one cohort. Outcome and group fields are complete in this example. The figures illustrate the display, not a customer result.
Program completion · illustrative cohort
150 eligible · 150 known outcomes · 150 usable group records
Group A: 60% · 60 of 100 completed
Group B: 80% · 40 of 50 completed
Overall: 66.7% · 100 of 150 completed
Difference: Group A is 20 percentage points below Group B.
Inspect the calculation and limits
Group A: 60 ÷ 100. Group B: 40 ÷ 50. Combined: (60 + 40) ÷ (100 + 50). The overall rate is not the simple average of 60% and 80%. The difference does not establish its cause or a legal conclusion.
Next review: Check access and delivery conditions with relevant participants and staff. Owner: program review lead. No cause assigned yet.
Both bars use the same zero-to-100% scale. Labels carry the information without relying on color. The overall rate remains visible, but it does not replace the group comparison.
Do not label the lower bar “failing group.” The measure describes a result under particular conditions, not the value or capability of the people in that group.
Read the gap before ranking the action
The widest numerical gap is not automatically the first intervention. Review the severity of the issue, the quality of evidence, the size and uncertainty of the comparison, the affected perspectives and the action within the team's control.
A smaller gap may concern a serious access barrier. A large gap may reflect incomparable definitions or different program mixes. Neither should be dismissed, but they call for different next steps.
Equal rates do not establish that a process is fair. A complete review may need to examine who was eligible, who heard about the program, who could enter and what happened after completion.
Prepare the data behind the view
- Define the unit. Decide whether the measure concerns people, households, organizations or sites.
- Define eligibility and time. Use the same inclusion rules for the numerator and denominator.
- Resolve known data issues. Review duplicates, corrections, missing outcomes and inconsistent periods.
- Document group information. Retain unknown or declined values and explain the coverage.
- Record the version. Keep the measure, source method and instrument version used for each period.
A persistent person identifier can help some longitudinal analyses, but it is not a universal requirement for an equity dashboard. Appropriate aggregate sources and anonymous surveys can be valid. The view must reflect the limits of the source.
Display missingness as part of the result
If some outcomes are unknown, show that alongside the comparison. Do not quietly count unknowns as unsuccessful or omit them from a headline without explanation.
Likewise, if group information is available for only part of the dataset, distinguish overall outcome coverage from group-analysis coverage. A dashboard can show a complete overall count while correctly limiting what it says about differences between groups.
When comparing periods, note changes in respondent composition and collection method. An apparent improvement may partly reflect who answered. AAPOR's survey guidance provides context for reviewing coverage and the design of survey-based measures.
Use local context without losing a shared view
Across schools, chapters or delivery sites, agree the limited shared core needed for comparison. Let local teams retain additional questions that help them run their work. Put definitions and known differences in a common dictionary.
If one program measures attendance and another measures demonstrated skill, do not aggregate them under “success.” If completion criteria vary, display the difference or establish a justified common measure before combining rates.
For an education-specific application, see equity and access in education. OECD's monitoring guidance discusses examining participation, progression and outcomes in the context of equity and inclusion.
Bring relevant accounts into the review
Qualitative evidence can help a team understand how people experience access, support or delivery. Show approved themes and a clear description of who contributed, not a claim that one quotation explains an entire group's rate.
Use human-owned theme definitions, inspect uncertain cases and retain contradictory accounts. When the codebook changes, preserve the version and review the revised analysis. Avoid mixing themes created under different definitions without explanation.
Sopact's approach connects this review to the relevant numbers and context, reducing repeated coding and manual joining. The qualitative and quantitative analysis guide includes the workflow visual and an illustrative effort comparison. That efficiency does not establish causality or remove review responsibility.
Make source access appropriate to the audience
Public, leadership and specialist review views may need different detail. An approved aggregate, a definition sheet and an audit trail can support trust without exposing names or raw sensitive comments to every reader.
Test small-group filters, overlapping views, quotations and exports together. A hidden row can sometimes be reconstructed by subtracting visible values from a total. Disclosure controls need to address the actual combinations available, not only one chart in isolation.
Use clearly labelled unavailable or withheld states. A suppressed value is not zero. Avoid graphics that imply its size when the number has intentionally been withheld.
What to evaluate in the workflow
BI platforms and spreadsheets can display valid disaggregated measures and link supporting material. The practical evaluation is how much work your team needs to keep collection, definitions, permissions, analysis and reporting aligned over repeated cycles.
With Sopact, assess the connected workflow using one realistic comparison. Include a changed definition, corrected record, missing outcome, locally different question and restricted source. Ask staff to reproduce the result and explain what the evidence does and does not support.
Inspect whether routine maintenance is manageable without repeatedly rebuilding joins and reports. Keep specialist statistical or evaluation methods where the question requires them. A dashboard is a review surface, not a substitute for those methods.
Return a finding with an action and a limit
A useful review note states the finding, relevant coverage, possible explanations to examine, the chosen next step, an owner and a review point. If no action is justified yet, say what additional evidence is needed.
Build the practice through the Impact Measurement & Reporting course. Use the report-writing guide and report examples to communicate the result and uncertainty clearly.
Watch: connect dashboard measures and stakeholder evidence
This companion video discusses evidence behind the dashboard. Use it with the interpretation and disclosure boundaries described above.
Frequently asked questions
What should an equity dashboard show?
Show the scope, relevant counts and rates, group comparisons, missingness, definitions and review actions. Include context and appropriate evidence access so the display supports interpretation.
Does a dashboard explain the cause of a gap?
No. It can make a pattern visible and help reviewers inspect relevant evidence. Determining why the pattern exists requires further investigation and suitable methods.
Should we always act on the widest gap?
Not automatically. Consider severity, evidence quality, uncertainty, affected perspectives and the practical action available. A numerical ranking alone is not a decision rule.
Must every dashboard reader see personal records?
No. Use access appropriate to the purpose. Approved aggregates, definitions and source methods can support review without exposing identifiable information.
Can we use existing BI tools?
Yes. Evaluate how the full workflow maintains collection, definitions, analysis and permissions. Sopact's connected approach may reduce repeated assembly while complementing specialist tools where needed.
Choose your next step
King Center
Related practice: reviewing participant feedback. The dashboard and figures shown here are fictional, not customer results or evidence of a causal equity effect.
Read the customer story →