Delphi Method - iterative expert consensus for criterion importance elicitation

Overview

Delphi iterates expert ratings until consensus is reached (CV < threshold, typically 0.20). The final weights are normalised means of expert ratings. Anonymity and controlled feedback prevent dominant-expert bias. Typically 2-4 rounds suffice.

Output

Weight, higher is better

Data

Crisp, expert input required

Weights

Derived internally, no weight source needed

Size

2+ alternatives, 3-12 criteria works best

Used for

Expert-driven decision making, MAGDM

How it works

1. 1

   Round 1-k: collect importance ratings from each expert; compute mean μ_j and CV_j = σ_j/μ_j per criterion. If CV_j < threshold for all j → stop; otherwise feed back statistics and collect round k+1.

   $${\mu }_j=\frac{1}{k}\sum _e{r}_{ej};C{V}_j=\frac{{\sigma }_j}{{\mu }_j};w_j=\frac{{\mu }_j}{{\sum }_l{\mu }_l}$$

   Dalkey & Helmer 1963, p.460 (consensus procedure; pending PDF page verification)

Fits when / Look elsewhere when

Assumptions to verify

• Domain experts available
• Experts can express consistent comparisons

Edge cases and pitfalls

• •If CV_j < threshold for all j → stop; otherwise feed back statistics and collect round k+1.

• CV threshold too strict (< 0.10): may require many rounds and expert fatigue. Too loose (> 0.30): consensus not meaningful.

Works with

How to cite