Disaggregation Quality

AI Prompt Templates

Copy a prompt into Claude, ChatGPT, or Gemini. Paste your document at the bottom and run.

Paste a document and get a scored quality assessment with evidence and revision priorities.

6,069 characters
You are an expert M&E data analyst with experience in inclusive and intersectional measurement. Score the disaggregation quality of the document I will provide using the rubric below. The document may be a MEL plan, indicator framework, survey, evaluation report, donor progress report, monitoring brief, or any document where data should be disaggregated.

SCORING RUBRIC - Disaggregation Quality
Score each dimension 1-5 using these criteria:

DIMENSION 1: Disaggregation Plan Completeness
- Score 5: All four elements present. Standard disaggregations specified (sex and age at minimum where applicable to the population). Context-relevant disaggregations specified (disability, geography, vulnerability category, beneficiary type as relevant to the program). Disaggregation rationale documented (why these categories were chosen, why others were not). The plan covers all key indicators or data points rather than being selectively applied to a few.
- Score 4: At least three of four elements present. Standard and context-relevant disaggregations specified; rationale or coverage partial.
- Score 3: Sex and age disaggregation specified but context-relevant categories missing or thin. Rationale not documented. Selective application.
- Score 2: Disaggregation mentioned in passing. Categories not fully specified.
- Score 1: No disaggregation plan or sex-only by default.

DIMENSION 2: Method-Disaggregation Alignment
- Score 5: All four elements present. Sample size supports planned disaggregation (sub-group power adequate for the comparisons that matter). Data collection method captures the disaggregation variables (e.g., the survey actually collects sex, age, disability status). Sampling method does not undermine sub-group representation (e.g., not a convenience sample that excludes a key group). Categories are operationalized consistently across instruments (disability defined the same way in baseline and endline, in survey and KII guide).
- Score 4: At least three of four elements present. Sample and instruments capture disaggregation; consistency or sub-group power partial.
- Score 3: Disaggregation variables collected but sample undersized for sub-group analysis. Definitions inconsistent across instruments.
- Score 2: Plan calls for disaggregation but instruments do not collect the variables. Sub-groups too small.
- Score 1: No alignment between disaggregation plan and data collection method.

DIMENSION 3: Analysis Disaggregation
- Score 5: All four elements present. The analysis plan specifies disaggregated analyses, not just aggregated totals. Sub-group differences are tested for significance where appropriate (or qualitative pattern comparison documented for non-statistical work). Intersectional disaggregation is conducted where relevant (e.g., women with disabilities, not only women OR only persons with disabilities). Small sub-groups are handled transparently (combined with rationale, footnoted with caveats, or analyzed with explicit caution).
- Score 4: At least three of four elements present. Disaggregated analysis specified; intersectional or small-group handling partial.
- Score 3: Disaggregated tables produced but no significance testing or pattern interpretation. No intersectional analysis. Small sub-groups dropped without explanation.
- Score 2: Aggregate analysis only with sub-group totals appended. No comparison logic.
- Score 1: Analysis is fully aggregated. Disaggregation collected but not analyzed.

DIMENSION 4: Reporting Disaggregation
- Score 5: All four elements present. Disaggregated data is reported in the main text and tables, not buried in an annex. Visualizations show sub-group differences clearly (e.g., grouped bars, faceted charts, comparison tables). Patterns across disaggregations are described in narrative, not only shown in a table. Where data cannot be disaggregated for a planned dimension, the limitation is explicitly stated.
- Score 4: At least three of four elements present. Disaggregation visible in main text; visualization or narrative interpretation partial.
- Score 3: Disaggregated data appears in annex tables only. Visualizations show aggregate. Narrative ignores sub-group patterns.
- Score 2: Disaggregated data referenced in passing. No tables or charts surface sub-groups.
- Score 1: Reporting is fully aggregated. No disaggregation visible to the reader.

DIMENSION 5: Use of Disaggregated Data
- Score 5: All four elements present. Findings reference disaggregated patterns, not only aggregate trends ("uptake rose to 62 percent overall, but only to 38 percent among women with disabilities"). Recommendations respond to disaggregated findings, with different actions for different groups where warranted. Decisions documented in the report or management response reference disaggregated evidence. Equity gaps surfaced by the disaggregation are addressed in next steps, not noted and ignored.
- Score 4: At least three of four elements present. Findings reference disaggregation; recommendations or decision use partial.
- Score 3: Findings mention sub-group differences but recommendations are generic. No equity gap follow-through.
- Score 2: Disaggregated data presented but not interpreted in findings or recommendations.
- Score 1: Disaggregation has no influence on findings, recommendations, or decisions.

OUTPUT FORMAT:
Return your assessment as a table followed by a summary:

| Dimension | Score (1-5) | Evidence from Document | Priority Revision |
|-----------|-------------|------------------------|-------------------|
| Disaggregation Plan Completeness | | | |
| Method-Disaggregation Alignment | | | |
| Analysis Disaggregation | | | |
| Reporting Disaggregation | | | |
| Use of Disaggregated Data | | | |

**Total: X/25**
**Band:** Strong (22-25) / Adequate (17-21) / Needs Revision (11-16) / Substantial Revision (5-10)
**Single Most Important Revision:** [One specific sentence]

For any dimension scored 1 or 2, add a brief explanation and a concrete revision example.

DOCUMENT TO SCORE:
[Paste your document or relevant sections here]

Scoring Criteria

Disaggregation Plan Completeness
5Excellent

All four elements present. Standard disaggregations specified. Context-relevant categories added with rationale. Plan covers all key indicators.

4Good

At least three of four elements present. Standard and context-relevant categories specified; rationale or coverage partial.

3Adequate

Sex and age specified but context-relevant categories thin. Rationale not documented. Selective application.

2Needs Improvement

Disaggregation mentioned in passing. Categories not fully specified.

1Inadequate

No disaggregation plan or sex-only by default.

Method-Disaggregation Alignment
5Excellent

All four elements present. Sample sized for sub-group power. Instruments capture disaggregation variables. Sampling preserves sub-group representation. Definitions consistent across instruments.

4Good

At least three elements. Sample and instruments capture disaggregation; consistency or sub-group power partial.

3Adequate

Variables collected but sample undersized. Definitions inconsistent across instruments.

2Needs Improvement

Plan calls for disaggregation but instruments do not collect variables. Sub-groups too small.

1Inadequate

No alignment between plan and method.

Analysis Disaggregation
5Excellent

All four elements present. Disaggregated analyses specified. Sub-group differences tested. Intersectional analysis where relevant. Small sub-groups handled transparently.

4Good

At least three elements. Disaggregated analysis specified; intersectional or small-group handling partial.

3Adequate

Disaggregated tables produced but no testing or pattern interpretation. No intersectional analysis. Small sub-groups dropped without explanation.

2Needs Improvement

Aggregate analysis with sub-group totals appended. No comparison logic.

1Inadequate

Analysis fully aggregated. Disaggregation collected but not analyzed.

Reporting Disaggregation
5Excellent

All four elements present. Disaggregated data in main text and tables. Visualizations surface sub-group differences. Patterns described in narrative. Limitations stated where disaggregation not feasible.

4Good

At least three elements. Disaggregation visible in main text; visualization or narrative partial.

3Adequate

Disaggregation in annex tables only. Visualizations show aggregate. Narrative ignores sub-groups.

2Needs Improvement

Disaggregation referenced in passing. No tables or charts surface sub-groups.

1Inadequate

Reporting fully aggregated. No disaggregation visible.

Use of Disaggregated Data
5Excellent

All four elements present. Findings reference disaggregated patterns. Recommendations respond to sub-group findings. Decisions reference disaggregated evidence. Equity gaps addressed in next steps.

4Good

At least three elements. Findings reference disaggregation; recommendations or decision use partial.

3Adequate

Findings mention sub-group differences but recommendations generic. No equity gap follow-through.

2Needs Improvement

Disaggregated data presented but not interpreted in findings or recommendations.

1Inadequate

Disaggregation has no influence on findings, recommendations, or decisions.

Score Interpretation

Total (out of 25)BandNext Step
22-25StrongDisaggregation is robust across the full data pipeline and equity-relevant. Use as-is.
17-21AdequateMost likely fix: extend disaggregation through analysis and reporting, not just collection. Address flagged dimensions before fielding or publication.
11-16Needs RevisionSubstantial revision required. The disaggregation plan exists but does not flow through to use. Use the Revise prompt to repair the analysis-to-use chain.
5-10Substantial RevisionDisaggregation is aspirational or sex-only. Rebuild starting from an intersectional plan that ties categories to decisions and equity questions.