4.2 Overview of Research Methods Adopted in this Study
4.2.2 Delphi Study
4.2.2.1 Concept of Delphi Method
The history of the Delphi method goes back to the 1950s (Okoli and Pawlowski, 2004).
This method was introduced by the RAND Corporation (Dalkey and Helmer, 1963; Okoli and Pawlowski, 2004) who conducted a project, Project DELPHI, as a study using expert opinions. The purpose of introducing this technique was to obtain the most reliable consensus from a group of experts about a specific subject under research (Dalkey and Helmer, 1963). Issues such as identification/prioritization and forecasting represent the first application of the Delphi method and the development of concept/framework represents the second type of application of this method (Okoli and Pawlowski, 2004).
The Delphi method has been identified by Duan et al. (2010) as seeking “to obtain consensus on the opinions of experts through a series of questionnaires that collect and
aggregate informed judgments on specific questions or issues”. Essentially, the Delphi method or technique employs a series of connected questionnaires (Brancheau et al., 1996). This method, as defined by Keil et al. (2002), involves only a limited sample, and it does not lend itself to a complex quantitative analysis; it is a structured and iterative process which aims to accomplish group consensus among a panel of selected experts in the field.
The main characteristic that distinguishes the Delphi method from other research methods is the type of respondents who participate in the study (Okoli and Pawlowski, 2004). This has been highlighted by Okoli and Pawlowski (2004), “Researchers have applied the Delphi method to a wide variety of situations as a tool for expert problem solving”. One situation that has been widely used is the ‘rank-type’ Delphi (Okoli and Pawlowski, 2004). In this type of situation, the Delphi method develops group consensus regarding the relative importance of issues (Okoli and Pawlowski, 2004). It is worth mentioning that Schmidt (1997) provides a detailed explanation on how to conduct this type of Delphi method, including guidelines on how to collect data, how to analyse them, and how to report the results.
During the Delphi process, a panel of experts list is created, and each expert is requested to respond to a particular question or task (Shan et al., 2010). The Delphi method enables researchers to gather and aggregate expert responses and then reveal these responses to them all (Shan et al., 2010). Then, the experts will be able to know what others in their group suggest and they might change their responses (Shan et al., 2010).
This occurs in a loop in which successive rounds of questionnaires provide a summary of the experts’ responses to the preceding questionnaire and ask the experts to re-evaluate their opinions in each round based on the summarized results given to them (Brancheau et al., 1996; Dalkey and Helmer, 1963). Questionnaire rounds are usually continued until an acceptable level of consensus is accomplished (Brancheau et al., 1996; Shan et al., 2010).
4.2.2.2 Using the Delphi Method in IS and e-Government Research
The Delphi method, like any other research method, has been used widely in the field of information systems and its applications. Okoli and Pawlowski (2004) provide examples of studies that have used the Delphi method in information systems research.
4.2.2.3 Rationale of Delphi study
In the context of this research, conducting a Delphi survey is a necessity. The reasons behind this decision are as follows:
1. After reviewing literature and conducting the exploratory study, the theoretical framework was developed and one of the dimensions was personal values of (Schwartz, 1992). However, when reviewing the literature, no studies investigated what personal values are relevant to e-government portal success.
2. Since the proposed framework has 13 dimensions including personal values, the proposed framework has many measurement items which will be very long to answer and this may make potential participants reluctant to complete it. Therefore, it is necessary to find out from the ten personal values those most relevant to e-government portals’ success. The irrelevant personal values will be neglected. This will help in shortening the survey questionnaire administered in the next phase of this research.
3. In general, this Delphi study is in line with the study of Haag et al. (2009). The authors of Haag et al. (2009)’ study decided to conduct it because no previous research had informed what personal values are relevant to online learning.
4.2.2.4 Identification of Experts
In fact, to conduct a successful Delphi study, it is essential to select qualified experts with a deep understanding about the subject (Okoli and Pawlowski, 2004). All experts in this Delphi study were people who had been actively involved in research in the following fields: e-government, website evaluation, culture and personal values and they have many publications in their specialized areas of research. Many of those experts, who were invited to participate, are editorial board members of one of the respected journals in the aforementioned fields.
4.2.2.5 Delphi Survey Administration
According to Duan et al. (2010), “When considering the level of consensus to be required, two or three rounds are normally preferred”. This Delphi study was conducted in two rounds. The controlled feedback from round two resulted in 4 value types receiving between 51% and 71% agreement and therefore, a further round was considered unnecessary. In (Duan et al., 2010) study, only two rounds were conducted after the selected factors by the experts received between 53% and 88% agreement in the second round.
Data was collected for the two rounds between December 2012 and March 2013. The experts who agreed to participate were 11 experts out of 40 invited for the first round.
Those 11 experts participated in the second round as well.
4.2.2.5.1 First Round
The data collected for the first round of this Delphi survey was mixed: quantitative and qualitative. The experts were sent, via email: the invitation to participate in the study; the link for the online questionnaire and the questionnaire in Microsoft Word file format were attached to the email. This gives the expert the option to either fill in the questionnaire in Microsoft Word file format or fill it in online.
The questionnaire in this round has closed-ended questions and open-ended questions.
It is divided into 5 main sections. Section one provides information on how to participate in this Delphi study. Section two provides detailed information about the study. Section three requires some basic information about the expert as a participant in the study.
Section four requires the expert to tick no more than five value types which should be particularly relevant or may have significant impact on e-government portal success.
Section five provides detailed explanation about the value types under investigation and which have been obtained from (Schwartz, 1992) and (Changingminds.org, 2012).
4.2.2.5.2 Second Round
The data collected for the second round of this Delphi study was quantitative only. The 11 experts were sent, via email, the invitation to participate in the second round; the aggregated result of the first round, the link for the online questionnaire and the questionnaire in Microsoft Word file format were included in the email.
In the second round, the number of instances that fell into each category and the descending order of frequencies of value types were sent back to the experts, who were also sent their previous answers in a customized invitation for each expert. The experts were requested to rank no more than 5 value types (1-5), where 1 = least relevant, and 5 most relevant.
The questionnaire in this round has closed-ended questions. It was customized to each expert who participated in the first round of this Delphi study. This customization includes the invitation of the expert by his/her name and including his/her selection of value types in the first round. The questionnaire is divided into two main sections. Section one compiles responses and presents them with the summary results which include responses of each expert. Section two provides detailed information about the study and provides detailed explanation about the value types under investigation which have been obtained from (Schwartz, 1992) and (Changingminds.org, 2012). This information can be used as reference when they need explanation of the terms used in the questionnaire.