Delphi Technique

Basic Principles and Procedures

The basic premise of the Delphi technique is to develop consensus among expert opinions on a particular subject. Although there may be variations in design, the data gathering process in general includes four phases: (1) expert panel members are selected to respond to an open-ended questionnaire, informed by extensive literature review, to gather their opinions about a specific topic or area of focus, (2) the input from each content expert is recorded to grasp group perceptions about the topic, (3) researchers further investigate expert views via a follow-up survey, and (4) researchers review all information after the experts have analyzed preliminary data and provided input. The time span between waves can range from 2 weeks to 1 month, depending on the number of statements provided in the initial review document.

The Delphi technique is most useful to gather opinions of experts who are regionally or geographically apart. The panel can range from a handful to over 100 people, depending on how many people are deemed to have expertise on the subject. Delphi studies usually comprise three to five waves of data collection; however, the number may change based on the subject matter being studied. The data analyses for each wave can vary depending on the researchers’ design and aims. Preferred methods of analyses include mean, median, and mode (measures of central tendency); standard deviation; and interquartile range. After the first wave or round, researchers begin to rank the answers provided and develop a list of items that experts review and provide additional input about. Researchers prioritize these items and clearly state their rationale for doing so. A survey is developed that allows experts to continue to provide input, modify their judgment, and indicate importance of each aspect of the topic. Subsequent waves provide strengthening of consensus about the subject matter.

The Delphi technique has been criticized for lacking rigor. Adaptations of the method are common, and its design has varied since the technique’s inception. To address this problem, in 2000, Felicity Hasson, Sinead Keeney, and Hugh Patrick McKenna published a set of guidelines for researchers aiming to use this method. They provide a detailed checklist researchers can use to guide their process and contend that the technique can be useful when attempting to gather opinions and develop consensus among a large group of individual experts.

...