Delphi survey questionnaire example: questions, rating scales, and feedback
Adapt an original example covering first-round questions, participant instructions, follow-up feedback, and consensus rules.
A questionnaire built around the decision
A Delphi survey questionnaire needs more than a list of statements. Participants must understand the judgment being requested, the rating scale, how feedback will be used, and what they will be asked to reconsider. The example below follows one item from an initial rating through feedback and a final classification.
The sample concerns an illustrative professional training framework. It is an original teaching example, not a validated instrument or a completed client study. Adapt the language and decision rules to your own research question.
Start with the decision the questionnaire must support
Our example asks: “Which capabilities should be included in a minimum training framework for people introducing a new workplace service?” The framework is the intended output. Importance for inclusion is the main rating dimension.
The first rating round starts with candidate statements developed through preparatory work. Because an initial list is supplied, describe how it was created rather than implying the panel generated every item. A classical design could instead begin with open-ended idea generation; modified Delphi designs can use a prepared list.
| Section | What to include |
|---|---|
| Introduction and consent | Purpose, participation expectations, privacy arrangements, and any study-specific consent requirements. |
| Relevant background | Only the participant characteristics needed to describe or compare the planned stakeholder groups. |
| Rating instructions | The population, context, time horizon where relevant, and the meaning of each scale band. |
| Candidate statements | One clearly defined idea per item, with a stable identifier. |
| Comments and omissions | Space to explain a judgment, suggest clearer wording, or identify missing content. |
| Next steps | When feedback will be available and whether another round is planned. |
Round 1: sample instructions and questions
Sample participant instructions
For each statement, rate how important the capability is for inclusion in the minimum training framework. Consider the role described in the study introduction. Rate importance even if the capability is currently difficult to teach.
Use 1–3 for low importance, 4–6 for moderate importance, and 7–9 for high importance. Select “unable to judge” if you do not have enough information or relevant experience to make a rating. Use the comment box to explain your reasoning or suggest wording changes.
In the next round, you will receive summarized panel feedback and have the opportunity to retain or revise your rating. Agreement with other participants is not required.
These are example anchors. A study about appropriateness, agreement, or likelihood needs different wording. The use of nine points does not, by itself, make a study a RAND/UCLA Appropriateness Method study.
| ID | Capability to rate for importance | Optional follow-up |
|---|---|---|
| T01 | Recognize a service request that needs escalation to a supervisor. | What information would make this statement clearer? |
| T02 | Explain the next step in the service process to a person using the service. | Which part of this capability matters most in practice? |
| T03 | Record a service exception using the agreed reporting process. | Are there contexts in which this capability is unnecessary? |
Ask about missing items
“What capability essential to the minimum framework is missing from this list? Please suggest one capability per response and explain why it should be considered.”
Decide how new suggestions will be screened, combined, and introduced to the panel. An item appearing for the first time in Round 2 has not received the same number of ratings as an original item.
Keep different criteria in separate questions
“How important and feasible is this capability?” asks for two judgments in one answer. If feasibility matters, add a separate question with its own anchors. An item may be highly important while requiring substantial resources to implement. Our Delphi criteria guide explores this distinction.
Round 2: show feedback that supports reconsideration
Keep the item ID and show whether its wording changed. Provide the participant’s previous rating where the design permits, the group distribution, and a balanced summary of relevant comments. Tell participants whether feedback is pooled or separated by stakeholder group.
Illustrative feedback for T01
Your previous rating: 6.
Numeric responses: 20.
Distribution: 2 participants selected 1–3; 3 selected 4–6; 15 selected 7–9.
Example reasons for high ratings: Participants considered timely escalation important when staff reach the limits of their authority.
Example concern: The statement may be difficult to interpret unless the escalation criteria are provided.
Follow-up question: After reviewing this feedback, how important is T01 for inclusion in the minimum framework? You may retain or revise your rating. Please explain any concern that remains.
The numbers and comment themes are fictional. In a real survey, show actual data and avoid selecting only comments that favor inclusion. A participant can understand the majority view and still reasonably disagree.
A worked consensus calculation
For illustration, suppose the protocol defines consensus for inclusion as at least 75% of valid numeric ratings in the 7–9 band and no more than 15% in the 1–3 band. This is a chosen example, not a universal Delphi rule.
- High-importance ratings: 15 ÷ 20 = 75%.
- Low-importance ratings: 2 ÷ 20 = 10%.
- The item meets both example conditions for inclusion.
If 22 people answered the item but two selected “unable to judge,” this example uses 20 numeric ratings as the denominator and separately reports the two nonnumeric responses. The protocol must specify that treatment beforehand. Also report missing answers and participation against the number invited; an item percentage alone does not describe retention.
If the rule requires agreement within each stakeholder group, repeat the calculation for each group. A pooled pass is insufficient. Distinguish a rule not met from a rule for exclusion: unless you predefined an exclusion rule, classify the remaining items as not reaching inclusion consensus or as requiring review. See consensus thresholds in Delphi studies.
Before launch, review the participant experience
Read the survey from the perspective of each stakeholder group. Check whether definitions are understandable, whether scales keep the same direction, and whether mobile layouts allow participants to see the question and choices together. Rehearse the feedback stage as well as the first questionnaire.
Methodological reviews describe substantial variation in Delphi design, including questionnaires and feedback. Document the choices that matter for your study rather than relying on the method name alone. Niederberger and Spranger’s overview of Delphi in health sciences.
Build the questionnaire in Surveylet
Surveylet supports configurable rating scales, comments, stakeholder groups, controlled feedback, and multi-round or Real-Time Delphi studies. A Real-Time Delphi design changes when participants receive feedback; it still needs clear instructions and a documented closing point.
You can provide your approved questionnaire for configuration through Complete Survey Care. Questionnaire development, methodological design, and a detailed final analysis require the appropriate separate service scope. Explore turn-key study support if you need help developing the study itself.
Plan your Delphi project
Tell us what your panel needs to evaluate and what you need to deliver. We can discuss the Surveylet workflow and support that fit your project.