ATTITUDES TOWARD AUTHORITIES AS A FACTOR OF THE POLITICAL TROLLING PERCEPTION ON SOCIAL MEDIA IN RUSSIA

Abstract

Political trolling has recently become a new manipulation tool employed in digital politics. A large body of political science research that addresses online trolling focuses however either on troll detection or the description of trolls’ activities and strategies. This literature however largely ignores the issue of the perception of trolls by social media users. As a result, scarce research sheds light on the effects of political trolling on social media. From a methodological perspective, the situation is exacerbated by the prevalence of studies that draw on troll detection via human annotation of social media accounts. Sidestepping the issue of troll perception can dramatically bias the results of empirical studies in this context. This paper helps bridge this gap in academic literature on political trolling by analyzing the perception of trolling on VK, one of most popular social media platforms in Russia. We draw on scholarship on selective perception to develop the hypothesis that people with different attitudes towards the Russian government will tend to label social media posts with the opposite stance as trolling. We test this hypothesis using an original collection of survey data that we analyze using state-of-the-art mixed effects regression models. Our analysis reveals that oppositionleaning respondents show differential perception of pro-government and anti-government posts and tend to label the former as trolling. However, we do not find similar results of pro-government social media users. Our findings indicate important methodological limitations of empirical studies based on human-labeled accounts and highlight starking differences in the perception of the online political information by governmentleaning and opposition-leaning respondents, thereby raising important questions for future research.