Academic journal article Electronic Journal of Business Research Methods

Delphi Method Variants in Information Systems Research: Taxonomy Development and Application

Academic journal article Electronic Journal of Business Research Methods

Delphi Method Variants in Information Systems Research: Taxonomy Development and Application

Article excerpt

Please note: An earlier and shorter version of this paper was published in PACIS 2016 proceedings.


Delphi is a method used to examine a complex problem through a group of experts. Experts are chosen as a data source in Delphi because of their special knowledge and experience regarding the issue under investigation. The experts provide data through questionnaires over several iterations. After each iteration, controlled feedback with the anonymized consolidated responses is provided to all participants. Consequently, experts can reflect and revise their opinions and judgements for the next iteration. The process stops when the research questions are answered. This may, for instance, be the case when consensus is reached, theoretical saturation is achieved, or sufficient information has been exchanged (Skulmoski, Hartman and Krahn, 2007; Linstone and Turoff, 1975; Delbecq, van de Ven and Gustafson, 1975).

The Delphi Method was first described in 1963 by Dalkey and Helmer, who conducted a Delphi study at the RAND corporation to apply expert opinions to a military problem. Over the years, Delphi has been applied in many research areas such as business, education, healthcare, and IS (Gupta and Clarke, 1996; Mitchell, 1991; Gallego and Bueno, 2014). The number of studies in the IS field using the Delphi Method is increasing (Gallego and Bueno, 2014), and Delphi appears to be an established method in IS research (Rowe and Wright, 1999; von der Gracht, 2012; Skulmoski, Hartman and Krahn, 2007; Gray and Hovav, 2008). Studies using the Delphi Method have, e.g., identified key IS management issues (Brancheau, Janz and Wetherbe, 1996), developed a descriptive framework of knowledge manipulation activities (Holsapple and Joshi, 2002), and investigated the IS outsourcing provider selection process (Chang et al., 2012). Although our paper focuses on IS research, we assume that similar observations and conclusions can be drawn in other disciplines that have adopted the method and are increasingly using it.

From a methodological perspective, researchers have proposed many variants of the Delphi Method. Main variants include Classical Delphi (Dalkey and Helmer, 1963), Decision Delphi (Rauch, 1979), Policy Delphi (Linstone and Turoff, 1975), and Ranking-Type Delphi (Delbecq, van de Ven and Gustafson, 1975; Schmidt, 1997). Furthermore, researchers have modified these main variants and proposed sub-variants (Chakravarti et al., 1998; Chang, 2006; Landeta and Barrutia, 2011; Tapio, 2003; Gupta and Clarke, 1996; Paré et al., 2013). While the method's modifiability can be considered as one of its advantages, "there is the danger that too much modification without ensuring rigor may threaten the validity of the original research approach" (McKenna, 1994, p. 1222), which may negatively impact its quality and credibility (Gupta and Clarke, 1996).

There are suggestions to improve the Delphi Method's rigor (e.g., Gallego and Bueno, 2014; Paré et al., 2013). However, these publications focus on improving the rigor of specific Delphi Method variants but do not contribute to clearing the ambiguity regarding the differentiation and definition of Delphi Method variants. The objective of this research is to address this gap and propose a taxonomy of Delphi Method variants. Thus, our study contributes to enhancing rigor in applying the Delphi Method in IS research. Our corresponding research questions are:

(RQ1) What Delphi Method variants are differentiated in IS research?

(RQ2) To what extent does a clear distinction exist between these variants?

(RQ3) What are clearly distinguishable Delphi Method variants and their characteristics?

(RQ4) How can a taxonomy be set up to clearly differentiate Delphi Method variants?

(RQ5) How can the taxonomy be applied to existing and new research to define Delphi Method variants purposefully and unambiguously?

The paper is organized as follows. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.