How to Conduct a Delphi Survey

Grant Faint/Getty Images

The Delphi survey research method is often used for forecasting, so it is not a stretch to see possibilities for new applications of the Delphi method within the realm of developing technology. The Delphi survey method relies on experts to moderate feedback throughout the process. Consumers are quite capable of assuming that role when it comes to using their interactive mobile devices. Consumers are refining their ideas about what they want from technology and are generally pleased to share their insights.

Mobile device applications also have the capability to embed Delphi-like surveys research in their platforms.

Learn More About Delphi Surveys Research

One of the key features of the Delphi survey research is that participants are unknown to each other, and responses are iterative in batches, thereby eliminating influence, in much the same way that the rules of brainstorming demand that participants withhold judgment. Delphi surveys can be done by e-mail, but more often than not, they are part of a survey research application that integrates the data, the summation and display, and the analysis at each stage.

Develop the Delphi Survey Research Questions

For an applied, rather than theoretical research application, the industry landscape, white papers, and the opinions of knowledge leaders, may replace a more conventional review of the literature. Conversations with customers through interactive marketing and social media can foster the development of relevant research questions.

Design the Delphi Survey Research

Typically, the Delphi survey method is used when a researcher is interested in collecting opinions from experts in a group setting, but also wishes to safeguard against undue influence of participants upon one another. Both qualitative and quantitative research methods can be used in a Delphi survey.

The critical consideration is which approach is most likely to answer the survey questions. As a final step, and outside of this basic outline, data from the Delphi survey process can be further validated by follow-up surveys.

Construct a Delphi Survey Research Sample

It is important that he survey participants have extensive experience or expertise in the area the research is investigating. Since the Delphi survey research is iterative by nature, it is also important that the participants have sufficient time to dedicate to completion of all the rounds of the survey. Conventional research considerations about representative samples and confidence intervals should be addressed.

Develop Round One of the Delphi Questionnaire

Questions in the first round of the Delphi survey are purposefully broad and invite participation that is in the spirit of brainstorming. It is important in this phase, as in all the other stages, to develop questions that are clear and are likely to result in equally clear responses.

Conduct a Pilot Study and Analyze Outcomes

In this stage, the researcher is given an opportunity to evaluate the scope of the research project. The researcher might be looking at how long it takes participants to answer the questions, and make any necessary adjustments that might either focus the questions more, or impose some limits on the length of acceptable responses.

The researcher and the panel of experts examine the questions again during this stage to see if they result in unclear responses.

Send Out and Analyze Round One Questionnaire

Final questionnaires are released to the survey participants in this stage. An analysis of the responses that are returned to the researcher is conducted once all the responses are back. Responses can be summarized and put into a map or heuristic that helps the respondents see patterns or understand the collective intelligence better in the next round. This is one of the reasons that a Delphi survey is moderated by experts. It often takes specialized knowledge to weed out responses that do not contribute to the conversation or are misleading.

Develop Round Two Delphi Questionnaire

Two of the goals of round two of the Delphi survey are to reduce the amount of information that has been shared by the respondents, and to develop the second set of questions based on the responses to the first set.

The researcher can wholly direct the next round of questions, allow himself or herself to be influenced by the responses received, or can combine both approaches.

Release and Analyze Round Two Questionnaire

As in round one, a set of questions is released to the study participants, and responses are analyzed. An additional task falls to respondents in this round -- validation that the summarized responses from round one do indeed reflect their intended meaning. Once the respondents have verified the relative accuracy of the summarized information, they turn to the questions that were generated by that information.

Develop Round Three Delphi Questionnaire

Once again, as in round two, the respondents are asked to answer questions that verify the summarization of their previous responses. Round three brings the researcher closer to understanding the limitations of the research and what research could potentially build on the findings.

Summarize and Document Delphi Survey Research Findings

In most Delphi surveys, this is the final stage. Respondents are given an opportunity to verify the collective responses, or to change their answers or comment further to clarify. If a sufficient amount of information has been collected and / or the research questions have been answered, and the respondents are in consensus, the Delphi survey process comes to an end. The process may also be terminated by the researcher or the experts if it is felt that it is not possible to further research a theory or that further questions will not produce new information.