# Methods Design Decision Memo — illustrative template

**Illustrative structure. No client information. Actual scope and evidence requirements vary by study.**

Published by Gurjas Evidence and Policy Analytics · gurjas.org
Template version 1.0 · 21 July 2026

---

## What this document is for

A methods decision memo records a single consequential design choice at the
point it is made — before the decisive result is known — so the reasoning
can be inspected later by a supervisor, reviewer, examiner or co-author. Use
one memo per decision, not one memo per study. A study typically produces
several: sampling frame, measurement model, handling of missing data,
choice of estimator, and so on.

Writing the memo before the result is the entire point. A decision explained
after seeing the data invites — and cannot rule out — the suspicion that the
explanation was reverse-engineered to fit a convenient result.

## Template

**1. Decision point**
State the single decision this memo covers, in one sentence.
*Example shape: "Whether to treat [construct] as reflective or formative in
the measurement model."*

**2. Context**
- Research question this decision serves:
- Stage of the study (design / data in hand / pre-analysis / other):
- Who needs to rely on this decision (supervisor, committee, journal reviewer, funder):

**3. Options considered**
List every option seriously considered, not only the one chosen.

| Option | Brief description | Main advantage | Main limitation |
|---|---|---|---|
| A | | | |
| B | | | |
| C | | | |

**4. Decision made**
State which option was selected.

**5. Rationale**
Explain why, referencing the research question and any governing literature,
guideline or prior registration — not the data itself if this memo is
written pre-analysis.

**6. Risks and trade-offs knowingly accepted**
State what this decision makes harder, weaker or impossible to claim later.
A decision with no listed trade-off has usually not been examined closely
enough.

**7. Reversal condition**
State what evidence, if it appeared, would cause this decision to be
revisited. A decision with no stated reversal condition is not falsifiable
in practice.

**8. Sign-off**

| Field | Value |
|---|---|
| Prepared by | |
| Reviewed by | |
| Date | |
| Revision | v1.0 |

---

*This is a structural template only. It contains no data, findings or
information from any Gurjas engagement. Adapt the sections to the scope of
your own study; do not treat the table headings as exhaustive or mandatory
for every decision type. Gurjas Evidence and Policy Analytics · gurjas.org*