Perceptions of the Australian public towards mobile internet e‑voting: risks, choice and trust

Abstract

This paper reports on data collected from an anonymous survey on perceptions of the Australian public towards using a mobile internet e‑voting platform (N = 295). It is the first such study conducted in an Australian context by an academic institution, which allows this research to be approached with a sense of impartiality. Our society has become rapidly fuelled by the mobilization of interactions and services. As the society becomes increasingly wirelessly connected, these mobile platforms are expected to provide an untapped universal medium by which paper based elections can be complemented or even "upgraded" to digital elections. This research is the first paper in a study which will be focusing on internet e‑voting, specifically the utilisation of mobility devices within Australia. As with any research, context shapes the direction and outcome goals. Internet e‑Voting (and research pertaining to) has gained momentum over recent years. Though there has been much research done in this field, there was been a gap in findings when dealing with Australian and mobility context, however similarities can be drawn from these related studies. The way the Australian context differentiates itself in one instance is Compulsory Voting. Utilising the findings from this initial study, we intend to provide a baseline from which our research can be further analysed and in turn will allow the derivation of hypotheses leading to creation of a user acceptance model towards utilisation of a mobile internet e‑voting platform during an Australian election. Survey respondents were overall more in favour of using mobile internet e‑voting (75.25%), with more respondents requiring greater information about the technology (15.93%) rather than being against its use (8.82%). The top appeals of the platform were towards mobility (91.40%), verifiability (72.90%) and speed (72.50%), with the top concerns towards manipulation (75.10%), retrieval (65.30%) and monitoring (63.20%) of casted votes by malicious parties or software. The initial hypothesis that were derived from the c