Introduction to Stakeholder Analysis

In Delphi studies, you rarely have a single, uniform group of participants. Instead, you invite experts from different backgrounds, such as:

• Industry leaders
• Clinicians
• Researchers
• Policy makers

These groups can have distinct perspectives, priorities, and experiences. Stakeholder Group Analysis helps you uncover and understand these differences systematically.

Why Do Stakeholder Analysis?

Consensus is rarely universal. One group might see a scenario as highly probable, while another dismisses it as unrealistic.

By analyzing subgroups, you can:

• Reveal hidden disagreements or alignments
• Understand why consensus fails in some areas
• Craft balanced recommendations that respect all perspectives
• Show decision-makers which stakeholders are more supportive or critical

How Stakeholder Group Analysis Works

Step 1: Define Your Groups

Before analysis, you decide which stakeholder categories matter.

You can segment participants by:

• Professional role (e.g., academic, practitioner, policy maker)
• Region (e.g., Europe, Asia)
• Experience level (e.g., junior, senior experts)
• Sector (e.g., public, private)

This step builds on your Delphi preparation: you already think of metadata when recruiting participants.

Step 2: Filter Ratings by Group

For each scenario or item durvey.org calculates and displays:

Compute group-specific statistics, for example:

• Mean ratings
• Consensus levels (e.g., interquartile range)
• Standard deviations

This helps you see whether groups agree or differ at a descriptive level.

Step 3: Compare Groups

You have two main ways to compare:

1. Basic (Descriptive) Comparison

Just compare the means, and consensus metrics side by side.

Example:

AI Diagnostics – Probability of Adoption

• Clinicians: Mean = 65%, IQR = 20%
• Tech Developers: Mean = 80%, IQR = 15%

Interpretation: Tech developers are more optimistic and show higher consensus.

This approach is simple and clear, but does not tell you if the difference is statistically significant.

2. Advanced (Statistical) Comparison

Use statistical tests such as the Mann-Whitney U-Test to check if differences are significant beyond random variation.

Example:

• Null hypothesis: Both groups rate AI diagnostics equally.
• Mann-Whitney U-Test result: p = 0.02

Interpretation: There is a statistically significant difference between clinicians and developers (p < 0.05). This adds rigor and credibility to your findings.

Example Scenario

Imagine you are researching emerging technologies in healthcare.

Research Question:

Do clinicians and tech developers agree on which innovations are most promising?

Findings:

• Clinicians rated AI diagnostics higher.
• Tech developers prioritized blockchain for data security.
• Mann-Whitney U-Test showed p < 0.05 for AI diagnostics.

Interpretation: Professional background clearly shapes expectations. For a successful strategy, you must address both groups' priorities. This is a classic example of why stakeholder group analysis is indispensable in interdisciplinary studies.

Why It Matters

Stakeholder Group Analysis helps you:

• Identify conflicting priorities early on.
• Communicate nuanced findings in your study.
• Develop balanced recommendations that integrate different perspectives.

Example Interpretation for a Report

"Clinicians rated AI diagnostics significantly higher in feasibility compared to tech developers (Mann-Whitney U = 42, p = 0.03). This highlights the importance of aligning technical capabilities with clinical acceptance."

How Durvey Helps

Durvey makes Stakeholder Group Analysis simple:

• Instantly filter and split your data by any participant attribute.
• Generate ready-to-use visuals and texts for your publications and presentations.