Action Plan — T-A-H Meeting 2026-06-06

Target: Research Policy Research Note — confirmed, stay the course.
Drop formative instrument framing — reframe entirely as composite index (Bloom & Van Reenen tradition). No more reflective CFA table in main text. No formative/reflective language.
Connection to RP: not through innovation outcomes (we don't have them) but through science itself — RP is about the science↔business interface; we measure whether managers use the scientific method. This is the Trojan horse for RP.
Style: econ-like, not OB. Brevity over thoroughness in text; details go to appendix.
New external validation: LLM annotation of RCT transcripts (see Task 4).
Template: Daniela Scur & Sarah Wolfolds "Revisiting the World Management Survey" — study the structure especially the back-end.

Action Items

Task 1 Rewrite framing — intro + conclusion Helena

New intro argument: science increasingly informs management practice → we need tools to measure whether individuals inside firms make decisions scientifically → SDM is that tool.

Key sentence: "Several decisions within firms still rest with an individual" — product management, project decisions, etc.
Individual-level variance > firm-level variance is the key empirical finding → reinforces the framing
Connection to RP in intro AND conclusion AND cover letter
Cut or rewrite any sentence that sounds like it belongs in OBHDP or JPSP

Task 2 Restructure main text (~5,000 words) Helena

Section	Words	Content
Introduction + background	~1,000	Science↔management framing, what we do, brief lit
Validation	~1,500	1 paragraph per test. Not CFA-heavy. Content validity, non-redundancy (VIF), robustness to alt weights, experimental sensitivity, nomological pattern.
Empirical patterns	~1,500	Show the distributions. Figures. Heterogeneity across firm tiers, industries, countries. The data IS the paper.
Back-end: implications for innovation research	~1,000	What can this data/tool be used for? Connect to RP. See WMS paper for template.

Move to appendix: CFA table, ICC details, acquiescence analysis, formative/reflective discussion, most robustness numbers.

Task 3 Write back-end section Helena

Read Scur & Wolfolds "Revisiting the World Management Survey." List 4–6 concrete research questions that SDM data opens up for innovation/RP readership. Frame as: "this tool is now available; here is what innovation researchers can do with it."

Task 4 LLM validation on RCT transcripts Helena + Arnaldo's method

Get transcripts of entrepreneurs from Diego's RCT data (treated vs. control). Convert 24 SDM items into annotation rubric → LLM scores each transcript → LLM-derived SDM score per person. Test: treated group score > control group score ("boiling water from another pot").

Method: robust annotation approach (Arnaldo, Abi, Said, Alfonso paper)
Backup for reviewers: persona generation from transcripts — leave for R&R

Blocking questions before starting:

Do we have access to Diego's transcripts? (Check with Arnaldo/Diego)
Which RCT? Which treated vs. control condition maps cleanest onto SDM?
IRB/consent covers secondary LLM use of interview transcripts?

Task 5 Circulate minutes + propose submission timeline Helena → Arnaldo + Thomas

Send this action plan to both. Propose timeline: estimate weeks for Tasks 1–4, then propose submission date.

What We Are Not Doing

No Prolific convergent battery (AOT/NFC) — not needed for RP Note
No Orbis linkage — not needed for RP Note; if we do it, it's a different paper
No additional survey data collection — too slow, too costly
No persona generation (LLM) — leave for reviewer response
No trying to be an OB/formative instrument paper

Open Items

Timeline

Week	Dates	What happens
Week 1	Jun 9–13	Request Diego's transcripts (Day 1). Read Scur & Wolfolds. Start rewriting: intro + framing + conclusion (Tasks 1, 3).
Week 2	Jun 16–20	Restructure main text, move things to appendix, finish back-end section (Tasks 2, 3). If transcripts arrived: start rubric + LLM annotation (Task 4).
Week 3	Jun 23–27	Finish LLM analysis if not done. Full pass clean draft. Internal review. Buffer for slippage.
Submission-ready	~Jun 29	Clean draft ready to share with Thomaz and Arnaldo before sending to RP.

Key dependency: transcripts from Diego gate Task 4 but NOT Tasks 1–3. Rewriting starts immediately regardless.

Target submission date: end of June 2026

Action Plan — T-A-H Meeting

Decisions

Action Items

What We Are Not Doing

Open Items

Timeline