Fix ContinuousDiD SE method: influence function, not delta method

[igerber]

Methodology references (required if estimator / math changes)

• Method name(s):
• Paper / source link(s):
• Any intentional deviations from the source (and why):

Validation

• Tests added/updated:
• Backtest / simulation / notebook evidence (if applicable):

Security / privacy

• Confirm no secrets/PII in this PR:

[github-actions]

Overall assessment: ⚠️ Needs changes

Executive Summary

• ContinuousDiD now documented as influence-function analytical SEs in docs/choosing_estimator.rst:L278-L280, which matches the registry and implementation.
• A conflicting statement remains in docs/methodology/continuous-did.md:L300-L302 that still says “delta method,” leaving mixed guidance for the same estimator.
• No code changes; resolving the documentation inconsistency should be sufficient.

Methodology

• [P1] ContinuousDiD SE method is now documented as influence-function-based in docs/choosing_estimator.rst:L278-L280, but docs/methodology/continuous-did.md:L300-L302 still states “delta method.” Impact: Conflicting guidance on analytical SEs vs. the Methodology Registry and implementation; risks users citing the wrong inference method. Fix: Update docs/methodology/continuous-did.md:L300-L302 to align with influence-function analytical SEs (or explicitly justify delta method if it applies), consistent with docs/methodology/REGISTRY.md:L453-L458 and diff_diff/continuous_did.py:L435-L446.

Code Quality

• No findings.

Performance

• No findings.

Maintainability

• No findings.

Tech Debt

• No findings.

Security

• No findings.

Documentation/Tests

• No additional findings beyond the Methodology note.
• Tests not run (doc-only change).

Open Questions

• None.

[igerber]

/ai-review

[github-actions]

🔁 AI review rerun (requested by @igerber)

Head SHA: 480fbe93cc7e3547a4d091fa678d195926fb6a16

---

Overall assessment: ✅ Looks good

Executive summary

• Docs now describe ContinuousDiD analytical SEs as influence-function-based, matching the Methodology Registry and current implementation.
• No estimator or variance code changes in this PR; behavior remains unchanged.
• No new edge cases or test gaps introduced by this docs-only update.

Methodology

• P3 (info) Impact: Documentation aligns with registry and implementation that ContinuousDiD analytical SEs use influence functions. Concrete fix: None. References: docs/choosing_estimator.rst:279, docs/methodology/continuous-did.md:300, docs/methodology/REGISTRY.md:457, diff_diff/continuous_did.py:435.

Code Quality

• P3 (info) Impact: Docs-only change; no code quality concerns in modified files. Concrete fix: None.

Performance

• P3 (info) Impact: Docs-only change; no runtime impact. Concrete fix: None.

Maintainability

• P3 (info) Impact: Docs update improves clarity without affecting maintainability. Concrete fix: None.

Tech Debt

• P3 (info) Impact: No new tech debt introduced by documentation edits. Concrete fix: None.

Security

• P3 (info) Impact: No secrets/PII introduced in documentation changes. Concrete fix: None.

Documentation/Tests

• P3 (info) Impact: Documentation updated appropriately; tests not required for docs-only change. Concrete fix: None.