Evaluation Plan

---
name: evaluation-plan
description: Draft the evaluation, measurement, data collection, or performance monitoring section of a grant proposal. Covers matching evaluation design to award size, selecting instruments and baselines, formative versus summative evaluation, data management and participant privacy, and reporting cadence.
---

# Evaluation Plan

## Overview

Reviewers use this section to test whether the outcomes promised earlier are
actually verifiable. An ambitious outcome with no credible measurement plan is
scored as an unsupported claim.

## When to Use

Drafting an evaluation plan, measurement strategy, data collection plan,
performance monitoring section, or logic model.

## Instructions

### Match the design to the award

Over-promising a randomized trial on a $40,000 grant damages credibility as
much as under-specifying does.

| Award size | Typical expectation |
|---|---|
| Under ~$100k | Pre/post measures, participation tracking, satisfaction data |
| ~$100k-$1M | Comparison group where feasible, validated instruments, external evaluator |
| Over ~$1M, research | Rigorous design, power analysis, IRB, data sharing plan |

### Required elements

1. **Evaluation questions** — mapped one-to-one onto the objectives. If an
   objective has no evaluation question, one of the two is wrong.
2. **Indicators** — for each question, the specific metric.
3. **Instruments** — named. Prefer validated instruments and cite them. Say who
   administers each and when.
4. **Baseline** — where each baseline value comes from. If none exists, say how
   and when it will be established.
5. **Data collection schedule** — what is collected, by whom, how often.
6. **Analysis** — how the data will be analyzed and what threshold counts as
   success.
7. **Use of findings** — formative feedback loops during the project, not just
   a summative report at the end. Name the decision points where data changes
   what you do.
8. **Reporting** — what the funder receives and when.

### Data management and privacy

State how participant data is stored, who can access it, how long it is
retained, and how it is de-identified. For research involving human subjects,
address IRB review explicitly. For minors, address consent and assent.

### Evaluator

Internal or external, named if known, with relevant qualifications and the
share of effort budgeted. An external evaluator listed here must appear in the
budget.

## Rules

- Every objective needs a matching evaluation question, and vice versa.
- Never claim a validated instrument without naming it.
- Targets must match the numbers used in the project design and budget exactly.
- Missing baselines are `[NEEDS INPUT: <question>]`, never estimated.
- Confirm length with `measure_text` before finishing.