M&E StudioTools

Draft, check and improve: Findings-to-Recommendations Quality

Draft, check and improve: Findings-to-Recommendations Quality: Your prompt

Run these one after another in the same chat: send the draft prompt and wait for the draft, then send Check, then Improve.

Writes the draft, with the criteria as the bar it must meet.

You are an expert M&E evaluator. Generate a findings-and-recommendations section that scores Excellent (5) on every dimension of the Findings-to-Recommendations Quality rubric for the situation described below.

QUALITY REQUIREMENTS:

Findings Specificity and Evidence Base: State concrete findings, not vague claims. Cite the specific evidence (data point, quote, observation) behind each finding. Indicate strength of evidence where it varies. Disaggregate findings by group, geography, or time where it matters to interpretation.

Evidence-to-Recommendation Linkage: Tie each recommendation to the specific finding or findings that generate it. Make linkage bidirectional so the reader can move both directions. Eliminate orphan recommendations and ensure every high-importance finding produces a recommendation where one is warranted.

Recommendation Specificity and Actionability: Write recommendations that name what to do, by whom, by when. Match the level of each recommendation to the decision authority of its target audience. Make recommendations operationalizable so a reader could act without further translation.

Prioritization and Feasibility: Rank recommendations or label priority. Consider resources, timing, capacity, and political feasibility. Address trade-offs where recommendations conflict. Note sequencing and dependencies.

Ownership and Accountability: Name the owner for each recommendation by role. Specify follow-up cadence. Document the tracking mechanism. Include a mechanism for revising recommendations as conditions change.

FORMAT: Present as a numbered section: (1) Key Findings with Evidence Base, (2) Findings-to-Recommendations Mapping Table, (3) Prioritized Recommendations with Owner / Timing / Feasibility, (4) Follow-up and Accountability Plan, (5) Revision Triggers.

SITUATION BRIEF:

[Paste the program context, findings summary, target audience for the recommendations, and any feasibility constraints here]

Quality bar: The result must meet these criteria. It will be reviewed against them afterwards.

  • Findings Specificity and Evidence Base: Whether findings are concrete rather than vague, each finding cites specific evidence supporting it, strength of evidence is indicated where it varies, and findings are disaggregated where relevant.
  • Evidence-to-Recommendation Linkage: Whether each recommendation traces to specific findings, the linkage is bidirectional, no orphan recommendations exist without an evidentiary basis, and no high-importance findings are left without recommendations.
  • Recommendation Specificity and Actionability: Whether recommendations are specific rather than vague, each names what to do and by whom and by when, the level matches decision authority, and recommendations are operationalizable without further translation.
  • Prioritization and Feasibility: Whether priorities are indicated among recommendations, feasibility is considered, trade-offs between recommendations are addressed, and sequencing or dependencies are noted.
  • Ownership and Accountability: Whether each recommendation has a named owner, follow-up cadence is specified, accountability mechanism is documented, and a mechanism exists for revising recommendations as conditions change.

Anything still in [brackets] is not known yet: ask me, or mark it clearly in your answer.

Rubric · Create a Document · about 435 words

Score how well findings translate into actionable recommendations across five dimensions using AI.

You'll Need

Your findings and recommendations section or full document

You Get Back

A draft, then a check, then an improved version, in one chat

Criteria