The Multiple Platforms Effect (MPE): A Quantification of How Exposure to Similarly Biased Content on Multiple Online Platforms Might Interact

Abstract

Over the past decade, controlled studies have identified nearly a dozen new forms of manipulation that can be used by search engines, social media platforms, microblogging platforms, and intelligent personal assistants. A recent study has shown that when users were exposed repeatedly to similarly biased content on the same platform, the net impact of those exposures was additive. We now ask the following: What happens when users are exposed to similarly biased content generated by different means on multiple online platforms? In the present experiment, which was randomized, controlled, counterbalanced, and double-blind, we exposed people to similarly biased content generated by different means on three different platforms – simulations of Google Search, Alexa, and X (f.k.a., Twitter) – presented successively and in random order, and we found that the impact of successive exposures was additive for both opinions and voting preferences pertaining to political candidates. Overall, the number of undecided voters voting for the favored candidate increased with each successive platform exposure by 42.4%, then 56.5%, then 66.7% over the pre-exposure level. We speculate that if Big Tech companies share values or political preferences, their net effect on our elections might be considerably greater than the effect of any individual company.

Keywords: multiple platforms effect, MPE, SEME, online manipulation, online influence, search engine manipulation effect