Can one large stakeholder group dominate a Delphi result?
An overall consensus percentage gives every participant equal weight. That can be appropriate—but when stakeholder groups differ greatly in size, the largest group may determine the pooled result even if smaller, essential groups disagree.
Overall consensus and stakeholder consensus answer different questions
A pooled analysis asks, “What proportion of all respondents supported this item?” A stakeholder analysis asks, “How did each relevant group respond?” Both can be useful. Problems arise when the study treats the pooled answer as the only answer even though the panel was intentionally designed to include distinct perspectives.
A simple sensitivity check
1. Pooled result
Calculate consensus across all participants. This reflects the panel as recruited and gives every respondent equal weight.
2. Group-specific result
Calculate the same measure within each stakeholder group. This reveals agreement that may exist in only one group or disagreement hidden by the total.
3. Equal-group result
As a sensitivity analysis, calculate each group’s result and then give the groups equal weight. Compare this with the pooled result to see whether group-size imbalance changes the conclusion.
How the interpretation can change
Conclusion remains the same
If pooled, group-specific, and equal-group results all support the same classification, confidence increases that the conclusion is not driven by one group’s size.
Conclusion changes
If the pooled result reaches consensus but the equal-group result does not, report the difference. The appropriate action may be revision, another round, a stakeholder-specific recommendation, or an explicitly qualified overall conclusion.
What to plan and report
Before the study
- Why each stakeholder group is needed
- Target recruitment by group
- Whether pooled consensus is the primary rule
- Whether every group must meet a threshold
- Whether an equal-group sensitivity analysis will be used
During and after the study
- Invitations, participation, and attrition by group
- Overall and group-specific distributions
- Any material difference between pooled and balanced results
- How qualitative comments explain group differences
- How disagreement affected item decisions
How Surveylet and Calibrum can help
Surveylet’s Stakeholder Analysis can organize and compare relevant participant groups so the overall result does not conceal important differences. Calibrum can help define the group structure, consensus rules, sensitivity analysis, attrition reporting, and final interpretation before data collection begins.
Recent research example
A 2026 international modified Delphi included persons with lived experience, emergency-care professionals, and academics. The researchers reported agreement overall and by group, and calculated a weighted mean that gave the three groups equal influence to assess whether unequal group sizes altered the findings. They also reported participation across three rounds and examined stability between rounds.
Reference: Blom-Ham WHW, Galenkamp N, Wolf LA, et al. Defining the concept of Mental Dysregulation in patients requiring ambulance and Emergency Department care: results of an international modified Delphi study. BMC Emergency Medicine. Published September 9, 2026.
Protect essential perspectives without hiding the pooled result
Tell us which stakeholder groups matter to your study and how the final decision will be used. We can help structure a transparent analysis that shows both the overall panel and the voices within it.