There are many I-voting approaches and solutions suggested in the literature (Krimmer et al., 2008). Each has its own advantages and disadvantages. This research attempts to assess the feasibility and acceptance of I-voting in Qatar by means of an experimental case study, a unique system was proposed for use.
The process of the experiment involved four phases to achieve the overall aim of the study:
1) Plan and design the experimental systems, 2) Test the system and experimental performance, 3) Analyse the results and discuss the findings,
4) Finally, compare and contrast the results and findings with the Estonia I-voting case study (see Figure 7.1).
Each phase is explained as following sub-sections.
Figure 7.1: Experimental process
7.2.1 Plan and design
The I-voting application design used in this experiment was based upon best-practice, and made in accordance with the Qatar election .. Considerations have been given to fulfilment of voting principles and their applicability to the situation in Qatar.
A face-to-face semi-structured interview was held with Abdul Rahman Al-Sulati, the head of the Permanent Election Committee in MOI, to discuss the process of replacing manual voting with I-voting. This interview included a discussion which identified the
Plan &
Design Test
Compare results and findings with
Estonia case study Analyse results & discuss findings ompare it with Run experiment
requirements for introducing a successful I-voting system in Qatar. Since the system is Internet-based, there is almost no limit to the number of personal computers which can be used to send information to the system.
The main outcome from the interview was to identify Qatar Election requirements for the voters and the system:
1. The original citizenship of voters must be Qatari or Qatari citizenship must have been granted at least 15 years earlier.
2. Voters must be aged 18 years or more.
3. Voters must not be convicted of dishonourable crime unless rehabilitated.
4. Voters must be resident in the electoral constituency in which electoral rights will be exercised.
5. Voters must not by employed by the armed forces or police. 6. Multicast voting must be forbidden.
7. The system must be easy to use and must have a Qatari „look and feel‟.
8. The system must satisfy basic voting principles such as security and anonymity.
A simple design was used, which required only a limited number of resources. Where votes are encrypted. This encryption system is based on private and public keys. A public key is given to the citizen in order to secure their data. The private key is used by the server to identify and confirm the eligibility of the voter while the design was simplified compared to other I-voting models, it allowed for a pilot study which determined the acceptance by the Qatari people of such a new technology. The algorithm was designed to ensure that the anonymity of the voters is maintained with the strictest confidence.
There were two experimental viewpoints:
(1) the voter environment, which allows authorised voters to cast their votes.
(2) the permanent election committee environment, which allows the committee to view the election results.
Chapter 7: Experimental Study: Comparison between Qatar and Estonia in I-voting
7.2.2 Test
After the design had been implemented and tested successfully by two computer experts from Almajaz Telecommunication Company, the testing and evaluation of the developed I-voting prototype was carried out by the researcher to assess its effectiveness in terms of fulfilment of voting principles (See section 1.1). Semi- structured interviews were held with the two computer experts in order to evaluate potential threats to the source code. They were asked questions oriented towards the technical aspects of the system and its ability to meet the voting requirements of Qatar. The experts were testing the functionality of the I-voting system based on the document provided by the researcher (See appendix C).
The experts believed that the system was sufficiently usable to act as a pilot study, providing a demonstration of how I-voting could take place in Qatar. They indicated that they thought it was well structured and easy to use, even by novice computer users. However, they said that it would only be useful on a small scale for demonstrating I- voting. This is due to the infrastructure employed to run the application which was only capable of handling an experiment and would not be practical for large-scale voting. They indicated that it would be unstable for larger scale application. Additional resources, servers and more advanced infrastructures would be required to introduce I- voting throughout Qatar for all its citizens. Also, the programming language (Visual Basic) used in the application was not capable of managing large scale operations and so they suggested using a more robust programming language, such as Java or PHP for a full scale implementation. Furthermore, the experts stated that the application required improvements to provide more authentication factors, such as the use of a smart card and biometrics. Also, the system needs to provide a method for distributing the required PIN for logging-in to the system, such as through SMS to ensure privacy. In addition, there is a need for advanced encryption method such as 3AES and a combination of mix-networking and blind signature to maintain anonymity. Additional suggestions were to provide a backup for recounting through paper-based voting and enabling a third party to monitor the integrity of voting processes.
The experts also suggested having an awareness programme and training sessions for people, especially those with little computer knowledge, to enable them to use the system easily. It was suggested that the media, academics and employers should play a role in the awareness programme to help introduce the society to such a new technology. The experts voiced the opinion that the I-voting system developed was a good step for Qatar since it was uniquely designed to fit Qatar‟s election requirements. However, they suggested a lot of improvements in the application are required before moving from the experimental stage to large-scale application (see Appendix B).
7.2.3 Running the experiment
After ensuring that the I-voting system was functioning properly, the experiment was carried out with a sample of 86 eligible Qatari voters from different demographic backgrounds (age, education and computer knowledge) selected by convenience from friends and relations of the researcher. Due to the large participant sample size, the experiment was held in 5 sessions each involve about 17 participants. This would ensure that all participants had adequate resources to participate in the experiment and the researcher had the capability to observe and assess their acceptance. The participants were motivated to join in the experiment by invitation letters which explained its significance and the importance of their participation. They were then given a consent sheet to sign, indicating their willingness to participate in the experiment and ensuring their confidentiality. Participants were given an oral short presentation which took ten minutes explaining about the I-voting prototype capability and how it worked (See Appendix B). The participants were asked to vote through the I-voting system within a maximum of 5 minutes including the time for authentication and vote casting (see Appendix B).
Later, the participants were asked to provide feedback on a questionnaire to evaluate and assess how acceptable they found the I-voting in practice and to determine the nature of any barriers to successful implementation that may exist. The questionnaire consisted of two sections, a background section and a post-ballot section. In the background section, the voters were asked to provide some background information about their gender, age, occupation, computer literacy, previous voting behaviour, and
Chapter 7: Experimental Study: Comparison between Qatar and Estonia in I-voting
their opinion of I-voting and the effects of ICTs on society. In the post-ballot section, specific questions were asked about the usability of the system, the quality of the system in terms of secrecy (privacy) and security (against fraud), about their viewpoints related to I-voting and some questions related to the voters‟ identity (see Appendix B).
An independent observer from the Qatari election committee (QEC) was involved at the end of each session to inspect the vote counting, where votes were counted manually and compared with the vote counting shown in the I-voting system in order to determine the accuracy of the I-voting counting process.
7.2.4 Analysis of results & discussion of findings
After the experiment, the questionnaire results were gathered and analysed quantitatively and qualitatively. Both the positive and negative reactions of voters were examined, as well as their ability to use the technology successfully. The results of this study indicated that the overall reaction of the voters towards the I-voting system was favourable. There was a high correlation between pre-voting anxiety and good perceptions of the new system. The vote casting stage had a slightly less positive reaction due to problems regarding the functionality of the system. There were difficulties with the pictures of the candidates, graphics describing how to use the new system, and explanations of encryption and decryption. The studied results indicated that voters reacted to the electronic system similarly to the traditional voting methods. However, individuals with high degrees of anxiety regarding the system often recorded a less positive experience with the new technology. The election committee showed their acceptance of the proposed system and of the accuracy of vote counting as result of comparing the manual count matched the electronic one.