What is Delphi Studio?

Delphi Studio is software for designing and running modified Delphi and other consensus studies. You use it to record the protocol, invite and enroll an expert panel, collect ratings and questionnaire answers, review the results between rounds, and keep the decisions.

The Delphi method is the research process. Panelists answer in private. The team returns an explicit summary of the group, not each person's answer. Panelists may answer again. The study states a stopping rule before collection starts. Agreement is one possible result. It is not guaranteed, and a study can also document persistent disagreement or uncertainty. What is the Delphi method? explains that process. This page explains the software.

Creating a draft does not launch a study, open a round, or send an invitation. Start a new study creates the draft. Conducting the study still means finishing Setup, launching, inviting the panel, opening each round, and recording the decisions between rounds. The research team makes those decisions. The software applies the rules the team confirmed. It does not decide what a result means for the study.

What you get

Capability	Detail
Setup before launch	Ten Setup steps cover identity, method, eligibility, panel size and mix, profile questions, the item bank, rating and consensus rules, the round plan, consent and communications, and the analysis plan. Launch does not freeze every study rule. What launch freezes, what stays editable, and what a recorded amendment can still change are in Launch and amendments.
Consensus rules stated in advance	Each statement is classified by the rules the study set on Rating & consensus: the share of ratings in the agree or disagree zone, the spread, and a minimum number of responses. A new study starts from editable seeds. Those seeds are not a universal standard. The results report shows the response count, the unable-to-rate count, zone percentages, the median, and the interquartile range. Quartiles and the disagreement index are stored. Consensus analytics says where each figure is shown and where it is not.
Sealed results	Results are computed after collection closes and stored as sealed versions. A correction is a new version with a reason, never an edit of the stored version.
A recorded decision between rounds	While another round follows, a statement that is not waiting for a review gets one disposition: Advance (rated again as written), Revise (reworded, then rated again), or Hold (not rated again). A statement still waiting for a review decision has none yet. An uncertain or polarized statement is not automatically Revise. The rules send it on as written, the disposition is Advance, and the plan tile reads Rated again. On the last planned round, the Decision stage groups that same outcome under Final: no consensus. The shorter label No consensus is not that group. Final: no consensus is not the only last-round result. The five final-result names are on Key concepts. Interpreting results maps an uncertain result to Advance while another round remains. A person confirms the plan. An override needs a written reason, and the decision log records the confirmed plan.
Separated identities	Once a person is enrolled, ordinary study screens show a pseudonym such as "ABC-0004". Peers do not see one another's names. Revealing a name needs a written reason and is audited. That is not a claim that every file omits names and email addresses. Identity and pseudonyms says which downloads can contain them.
Simple panelist access	Panelists use a personal link from their email, with no account or password. Answers save as they go, and the study can offer "Unable to rate".
Files for a manuscript	Consensus, participation, and responses CSVs, a counts-only study summary, and a study audit CSV. These files support reporting. They are not a finished manuscript, and they do not by themselves meet a reporting guideline.

Who it's for

The documentation is written for research teams who design a consensus study and may need to explain the record to a reviewer or an ethics board. Examples include:

• Clinical guideline groups
• Medical education researchers
• Healthcare consensus panels
• Policy and governance working groups
• Research consortia

Each deployment serves one organization. Staff sign in with a link sent by email; there is no self sign-up.

How this documentation is organized

This page is the orientation. The staff guides are the procedures.

Section	Start here if you want to…
Your first study	Follow one new study from the draft through launch
Staff guides	Create a draft and learn each screen
Methodology	Learn the Delphi method and choose a design
Concepts	Understand studies, panels, rounds, and ratings
Using the platform	See how a study moves from Setup to completion
Analytics	Read the statistics and the classification
Reporting	Map the study record to ACCORD and CREDES
Worked examples	Follow end-to-end scenarios

Study structure

A study is built around a panel, an item bank of statements, a round plan, controlled feedback between rounds, and consensus rules the study states before it collects ratings. What those rules lock, and what can still change, is in Launch and amendments. Around those sit an organization-wide expert directory, reusable study templates, consent versions that are frozen once published, and withdrawal handling that keeps a person's submitted answers unless they separately ask for them to be removed.

See Structure for how these parts fit together.

Next steps

1. Your first study: the path for a new principal investigator. Creating the draft, launching, inviting, and opening a round stay separate.
2. Start a new study: create a draft and choose a workflow. This step does not launch the study or invite the panel.
3. What is the Delphi method?: the research process, separate from the software
4. Study design: the first group of Setup steps
5. Governed study record: access, saving, imports, and what launch changes
6. Roles and access: who on a study team can do what

note

This documentation describes Delphi Studio. To run a study, sign in at delphistudio.org.