Delphi Panel Size: How Many Experts Do You Actually Need?

No aspect of Delphi methodology generates more variation in published studies than panel size. The literature contains Delphi studies with 7 panelists and others with 170. Both can be methodologically defensible depending on the study's purpose, scope, and how expert is defined. This guide covers what the evidence actually supports — not a single correct number, but a framework for choosing the size that fits your study.

Why There Is No Universal Minimum or Maximum

Delphi panel size is not regulated by the same statistical logic that determines sample size in quantitative studies. In a survey study, sample size is driven by the need for statistical power relative to a target population. In a Delphi study, panelists are not a statistical sample — they are a purposive group recruited because they hold expert knowledge relevant to the question. The relevant criterion is representativeness of the expert opinion space, not statistical representativeness of a broader population.

This means the right panel size depends on how diverse expert opinion is in your domain, how homogeneous or heterogeneous your expert population is, and whether multiple stakeholder groups need representation. A panel of 10 highly heterogeneous experts can produce more informative consensus than a panel of 40 experts who all trained in the same tradition.

Evidence-Based Guidance on Panel Size

The Attrition Problem

One of the most reliable empirical findings in Delphi methodology is that panel dropout between rounds is common: rates of 10–25% attrition between rounds are typical in academic studies. Panelists who find the feedback from Round 1 insufficiently engaging or who feel their position is poorly represented are more likely not to return for Round 2.

The implication is simple: recruit with the buffer built in. If you want 15 active panelists in Round 2, you need to begin Round 1 with at least 18–20. If your panel has a subgroup you specifically need to retain — patients alongside clinicians, for example — track their attrition separately rather than treating dropout as random.

Defining Expert Before Setting Size

The most important decision in panel composition is not the number — it's the definition of expert. A clear, written expert definition does three things:

Expert definitions typically combine two or three criteria: a minimum level of experience (years in practice, seniority), a domain-specific credential or role, and sometimes a record of engagement with the specific question. All three in combination is stronger than any one alone.

Multidisciplinary Panels and Subgroup Balance

When a study requires multiple types of experts — common in healthcare, education policy, and sustainability research — panel size cannot be determined without first deciding the stakeholder structure. A panel with three distinct groups of roughly equal size behaves very differently from a panel where one group is dominant. Consensus reached in a majority-dominated panel may not reflect genuine multidisciplinary agreement — just the views of whoever had the most seats.

If your study requires stakeholder balance, set minimum sizes for each group rather than a single overall target, and report the composition breakdown in your methods section.

What to Write in Your Methods Chapter

For methods chapters and journal submissions, panel size should be reported with justification, not just stated. The expected elements are:

A methods section that reports only "N=14 experts participated" without explaining why 14, who they were, and what dropout occurred is a common reason for major revisions in peer review.