Quickstart: your first study
This guide walks you through launching a minimal modified e-Delphi study in Delphi Studio. Expect 30–45 minutes for setup if you already have items drafted and a panel list ready.
Watch the walkthrough
The video covers the same path as the written steps below. Use whichever you prefer — the written guide is the reference for exact field names and thresholds.
Before you begin
| Prerequisite | Why it matters |
|---|---|
| Researcher account | Sign in at delphistudio.org |
| IRB approval or exemption (if required) | Captured on the Identity step; bound to consent |
| Draft items (5–15 statements) | Imported or entered in the item bank |
| Panel emails (10–20 experts) | Magic-link invitations; no accounts needed |
Start with the Modified e-Delphi preset unless you are running a RAND/UCLA appropriateness study. You can customize every default before launch.
The five-step path
Step 1 — Create a study
- From the Dashboard, click New study.
- Enter a title (e.g. Core competencies for rural telehealth nursing).
- Choose Modified e-Delphi as the starting method.
- Delphi Studio creates a draft study and seeds email templates.
Step 2 — Complete the design wizard
Work through the 11 wizard screens. For a first study, focus on these essentials:
| Screen | What to configure |
|---|---|
| Identity & IRB | PI, institution, IRB number, protocol version |
| Objective & scope | Research aim and the consensus question |
| Method | 2 planned rounds, anonymous panel, controlled feedback |
| Panel rules | Target size (e.g. 15), expertise fields for subgroup analysis |
| Item bank | Add or CSV-import your statements; assign domains |
| Rating & consensus | 5-point Likert, ≥80% agreement threshold, IQR ≤ 1 |
| Rounds & feedback | Round window (e.g. 10 days), reminder schedule |
| Consent & comms | Review consent text and invitation email |
| Analysis & reporting | Response-rate floor, export formats |
| Review & launch | Resolve blockers; acknowledge warnings |
See the full Study design wizard for every screen.
Step 3 — Launch the protocol
On Review & launch:
- Fix any blockers (missing items, unset thresholds, empty consent).
- Acknowledge warnings (e.g. small panel size) with a documented rationale.
- Click Mark protocol complete → Open recruitment → Launch Round 1.
At launch, Delphi Studio freezes your methodology into a launch snapshot. What you configured is what you will report.
Step 4 — Recruit the panel
- Go to Panel and add panelists by email.
- Send invitations — each panelist receives a magic link.
- Monitor states:
pending→invited→accepted. - Panelists complete informed consent before their first rating.
Panelists appear as pseudonyms (Panelist 1, Panelist 2, …) everywhere in the researcher UI.
Step 5 — Run Round 1
- Round 1 is created automatically at launch — open Rounds to find it (use Plan next round for later rounds).
- Set
opensAt/closesAt, then open the round. - Panelists rate items on any device; optional comments are collected.
- Close the round when the window ends (or close manually).
- Review per-item result cards on the Results page; the heatmap is on the Report page.
If items are indeterminate, revise wording, plan Round 2, lock feedback packets, and open the next round. Panelists then see controlled group feedback before re-rating.
What to do next
- Key concepts — studies, panels, rounds, ratings
- Running rounds — scheduling, grace windows, feedback lock, stopping rules
- Panel management — stakeholder balance, unavailable state, bulk CSV
- Interpreting results — heatmaps, re-analysis after withdrawal
- Collaboration & roles — invite co-investigators
- Clinical guideline example — a full scenario with numbers
Delphi Studio runs background jobs (round auto-open/close, reminders, analytics) via a worker process. In production this runs continuously; your study operations do not require manual cron setup.