Examining the Internet Research Agency's Exploitation of Cognitive Biases Through its Disinformation Campaign Targeted at the US 2016 Presidential Election

Abstract

Numerous studies have been dedicated to the Kremlin-affiliated Internet Research Agency's (IRA) online disinformation effort dedicated at the US 2016 Presidential election. Research has confirmed that the effort involved a coordinated effort to influence the voting behaviour and sow discord amongst US citizens across social media networks. This paper differed from previous research in that it analysed the literature to evaluate if the IRA employed tactics designed to exploit specific cognitive processes including the vividness bias, availability heuristic, and confirmation bias. This research performs unique analysis for each bias by examining qualitatively how confirmation bias may have been exploited based on the IRA's network infiltration, performing quantitative analysis of the IRA tweet volume relative to a control dataset of humans and automated bot accounts during the pre-election period, and the vividness bias by employing a Sentiment Analysis to assess positive and negative sentiment in IRA tweets. Each bias is then evaluated in the context of Russian reflexive control doctrine and the greater IRA strategy. This paper finds that the IRA employed a savvy understanding of how cognition and social media network infrastructure influenced US users' behaviour and leveraged that knowledge to...