Real-Time Delphi vs. Multiple-Round Delphi
The Delphi method is a structured expert survey technique used to collect and refine opinions through controlled feedback. Two main variants are commonly used today.
Both approaches follow the same core idea: experts answer a set of questions, receive feedback about the group results, and may revise their answers. The goal is typically to reach consensus, identify disagreements, or prioritize future developments. However, the way feedback is provided and how the survey is conducted differs substantially.
Multiple-Round Delphi
The classical Delphi method is conducted in several separate rounds. Each round consists of a full survey, followed by analysis, feedback preparation, and a new round.
How It Works
- Experts answer the first questionnaire.
- Results are analyzed.
- Feedback (e.g., median, distribution, comments) is prepared.
- A new questionnaire is created including this feedback.
- Experts answer again.
- The process repeats until consensus or stability is reached.
This means the researcher manually controls each round.
Characteristics
- Several separate survey rounds
- Feedback is prepared between rounds
- Participants only see feedback in the next round
- More control for the researcher
- Longer duration
Typical Use Cases
- Academic studies
- Forecasting research
- Policy studies
- When careful iteration is required
- When qualitative comments are important
Real-Time Delphi
In a real-time Delphi, the feedback process is automated. Participants see the current group results immediately after answering a question and can revise their responses without waiting for another round. The survey runs continuously instead of in separate rounds.
How It Works
- Experts answer a question.
- The system immediately shows group statistics (e.g., median, mean, distribution, comments).
- Experts can change their answer at any time.
- The survey remains open until enough responses are collected.
There are no fixed rounds. The iteration happens automatically.
Characteristics
- Single continuous survey
- Instant feedback
- No manual round preparation
- Faster data collection
- Less administrative effort
- Ideal for online platforms
Typical Use Cases
- Large expert panels
- Time-critical studies
- Technology foresight
- Industry surveys
- Real-time consensus building
Key Differences
| Feature | Multiple-Round Delphi | Real-Time Delphi |
|---|---|---|
| Structure | Several separate rounds | One continuous survey |
| Feedback | Provided between rounds | Provided instantly |
| Effort | High | Low |
| Duration | Longer | Shorter |
| Control | High researcher control | More automated |
| Flexibility | High | Medium |
| Suitable for large panels | Limited | Very suitable |
Example
Research question: What technologies will have the biggest impact on the energy sector by 2040?
Multiple-Round Delphi
- Round 1Experts rate 20 technologies.
- AnalysisMedian and comments are calculated.
- Round 2Experts see the results and can revise their ratings.
- AnalysisSome items converge, others still differ.
- Round 3Experts rate again only the items without consensus.
- ResultConsensus ranking of technologies.
Duration: several weeks or months.
Real-Time Delphi
- Experts receive the survey link and rate the same 20 technologies.
- After answering each item, they immediately see:
- Median rating
- Distribution
- Comments from other experts
- They can immediately change their answer.
- The survey stays open for 2 weeks.
Result: Consensus emerges during the survey.
Duration: days or weeks.
When to Use Which Method
Use Multiple-Round Delphi when …
- •Careful iteration is needed
- •Feedback must be controlled
- •Panel is small
- •Study is academic
- •Qualitative input is important
Use Real-Time Delphi when …
- •Many experts participate
- •Fast results are needed
- •Survey is online
- •Feedback should be automatic
- •Administrative effort should be low
Summary
Both methods follow the same principle: structured expert feedback with iteration. The difference is how iteration is implemented:
- Multiple-Round Delphi → iteration between rounds
- Real-Time Delphi → iteration within one survey
Choosing the right variant depends on the research goal, time constraints, and panel size.