Trust, reputation and manipulation: a theoretical and empirical perspective applied to the case of eBay.

Summary To this day, trust has a polysemous character in social sciences. This notion becomes the center of interest of a number of sociologists, philosophers, psychologists and economists. The "six chapters" of this thesis address the issue of establishing trust between buyers and sellers during an online purchase. The reputation of a seller is presented as the main determinant of trust. The rating system adopted by the online sales site eBay is the focus of our empirical and theoretical analyses. These six chapters are preceded by a general introduction and a review of the literature (first chapter). A second chapter is dedicated to present the impact of ratings - especially negative ratings - left by buyers to sellers, on the latter's presence on the platform, using Markov Chain. A third chapter focuses on optimizing the sales of a specific seller during a given period, while taking into account the ratings left by buyers. The fourth is an empirical study that focuses on the impact of a seller's reputation on the probability of sale and on trust. Chapter five presents the limitations faced by the rating system and proposed solutions. The final chapter, measures and empirically tests the impact of seller reputation manipulation on sale probability and trust.