Sample Justification

Modèles de prompts IA

Copiez un prompt dans Claude, ChatGPT ou Gemini. Collez votre document en bas et exécutez.

Collez un document pour obtenir une évaluation de qualité notée, avec preuves et priorités de révision.

5,128 caractères
You are an expert M&E sampling specialist. Score the sample size and selection rationale section of the methodology I will provide using the rubric below.

SCORING RUBRIC - Sample Justification
Score each dimension 1-5 using these criteria:

DIMENSION 1: Analytic Purpose Match
- Score 5: All elements present. The sample size supports the intended overall analysis. Sample size supports planned disaggregation (e.g., by sex, age, region, treatment status) with adequate cell sizes. Sample size supports any planned subgroup comparisons. The rationale explicitly links sample size to each planned analysis rather than treating the overall total as sufficient.
- Score 4: At least three of four elements present. Overall and most disaggregation supported; one or two subgroup analyses thin.
- Score 3: At least two of four elements present. Overall sample justified but disaggregation thin or comparisons underpowered.
- Score 2: Sample size matches overall analytic purpose but cannot support stated disaggregation or comparison.
- Score 1: Absent or inadequate. Sample size cannot support the intended analyses.

DIMENSION 2: Statistical Justification
- Score 5: All elements present (where quant inference is intended). Confidence level stated (e.g., 95%). Margin of error or detectable effect size stated. Expected variability or baseline rate assumption stated. Design effect for clustered sampling stated. Non-response buffer applied. If no quant inference is intended, this is stated explicitly and the dimension is scored on completeness of that statement.
- Score 4: At least three of four elements present. Confidence and one of effect size or variability stated; design effect or buffer partial.
- Score 3: At least two of four elements present. Confidence and one assumption stated. Design effect and buffer missing.
- Score 2: Sample size stated for quant analysis but assumptions absent. Cannot be tied to power.
- Score 1: Absent or inadequate. No statistical justification offered for quant inference.

DIMENSION 3: Qualitative Saturation Logic
- Score 5: All elements present (where qual inference is intended). A target number is stated as a planning estimate. Saturation or information-power criteria are defined. The point at which saturation will be assessed is specified. The procedure for extending the sample if saturation is not reached is named. If no qual inference is intended, this is stated explicitly and scored on completeness of that statement.
- Score 4: At least three of four elements present. Target and saturation criteria stated; assessment point or extension procedure partial.
- Score 3: At least two of four elements present. Target stated but saturation criteria weak.
- Score 2: Target stated without saturation logic ("we will interview 30 people").
- Score 1: Absent or inadequate. No saturation or information-power logic provided for qual inference.

DIMENSION 4: Selection Method Fit
- Score 5: All elements present. Selection approach (probability, purposive, theoretical, mixed) matches the population. Approach matches the inference goal (generalization, depth, comparison). Selection criteria are operationalized. The approach is appropriate for the access conditions in the field.
- Score 4: At least three of four elements present. Approach matches population and inference goal; operationalization or access fit partial.
- Score 3: At least two of four elements present. Approach named and broadly fits; operationalization implicit.
- Score 2: Approach mismatched (e.g., purposive sampling proposed for population-level prevalence claims).
- Score 1: Absent or inadequate. No selection approach or approach clearly wrong for the question.

DIMENSION 5: Feasibility
- Score 5: All elements present. Sample size is achievable within the access conditions described. Achievable within the timeline. Achievable within the budget envelope. Field team size and capacity can deliver. Risks to feasibility (seasonality, security, mobility) are acknowledged and mitigated.
- Score 4: At least three of four elements present. Achievable on most dimensions; one feasibility risk partial.
- Score 3: At least two of four elements present. Plausibly achievable but feasibility risks not addressed.
- Score 2: Sample size implausible given access, time, or budget. Field plan does not support delivery.
- Score 1: Absent or inadequate. Sample size cannot be delivered as proposed.

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

| Dimension | Score (1-5) | Evidence | Priority Revision |
|-----------|-------------|----------|-------------------|
| Analytic Purpose Match | | | |
| Statistical Justification | | | |
| Qualitative Saturation Logic | | | |
| Selection Method Fit | | | |
| Feasibility | | | |

**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.

SAMPLE JUSTIFICATION TO SCORE:
[Paste your sample size and selection rationale here]

Scoring Criteria

Analytic Purpose Match
5Excellent

All elements present. Sample supports overall analysis, planned disaggregation (sex, age, region, treatment status) with adequate cell sizes, and any planned subgroup comparisons. Rationale links sample size to each planned analysis.

4Good

At least three of four elements present. Overall and most disaggregation supported; one or two subgroups thin.

3Adequate

At least two of four elements present. Overall justified but disaggregation thin or comparisons underpowered.

2Needs Improvement

Sample supports overall analysis but cannot support stated disaggregation or comparison.

1Inadequate

Absent or inadequate. Sample cannot support the intended analyses.

Statistical Justification
5Excellent

All elements present where quant inference is intended. Confidence level, effect size or margin of error, expected variability, design effect, and non-response buffer all stated. Where no quant inference is intended, this is stated explicitly.

4Good

At least three of four elements present. Confidence and one of effect size or variability stated; design effect or buffer partial.

3Adequate

At least two of four elements present. Confidence and one assumption stated. Design effect and buffer missing.

2Needs Improvement

Sample size stated for quant analysis but assumptions absent. Cannot be tied to power.

1Inadequate

Absent or inadequate. No statistical justification offered.

Qualitative Saturation Logic
5Excellent

All elements present where qual inference is intended. Target number stated as a planning estimate. Saturation or information-power criteria defined. Assessment point specified. Extension procedure named. Where no qual inference is intended, this is stated explicitly.

4Good

At least three of four elements present. Target and saturation criteria stated; assessment or extension partial.

3Adequate

At least two of four elements present. Target stated but saturation criteria weak.

2Needs Improvement

Target stated without saturation logic.

1Inadequate

Absent or inadequate. No saturation logic.

Selection Method Fit
5Excellent

All elements present. Approach matches population. Approach matches inference goal (generalization, depth, comparison). Selection criteria operationalized. Approach appropriate for access conditions.

4Good

At least three of four elements present. Match and inference goal clear; operationalization or access fit partial.

3Adequate

At least two of four elements present. Approach named and broadly fits; operationalization implicit.

2Needs Improvement

Approach mismatched to inference goal.

1Inadequate

Absent or inadequate. No approach or approach wrong for the question.

Feasibility
5Excellent

All elements present. Sample achievable in access, time, budget, and team capacity. Feasibility risks (seasonality, security, mobility) acknowledged and mitigated.

4Good

At least three of four elements present. Achievable on most dimensions; one risk partial.

3Adequate

At least two of four elements present. Plausibly achievable but risks not addressed.

2Needs Improvement

Sample implausible given access, time, or budget.

1Inadequate

Absent or inadequate. Sample cannot be delivered as proposed.

Score Interpretation

Total (out of 25)BandNext Step
22-25StrongSample justification is defensible. Minor refinements only.
17-21AdequateAddress flagged dimensions before fielding begins.
11-16Needs RevisionSubstantial revision required. Use the Revise prompt as a revision brief.
5-10Substantial RevisionSample will not support the intended analyses. Redesign before fielding.