Investigating Information Provenance as a Cue to Look through the Opacity of (Mis)Information

Abstract

Everyday users access information through different online media platforms as it propagates through the collective actions of users and algorithms. A diverse and huge audience with easy online access to such socio-technically curated information finds it challenging to make sense of this often evolving and sometimes contrasting information. Though researchers and some users know how to interpret this opacity of information, very few everyday users have the understanding and/or tools needed to look into online platforms’ machinations and witness how and why (mis)information spreads. How can we (re)design online media platforms to allow users to look through the opacity of propagating information — in often limited attention span of user interaction on these platforms — and assess the credibility of that information? This dissertation adopts information provenance — a record of information as it moves across users and platforms due to socio-technical actions — as a construct to ground thinking related to the overarching research question. Provenance facilitates a way to imagine what users could know about information and subsequently assess the credibility of the information. The first study on the Google Search platform suggested the importance of information provenance. Findings suggest that making this provenance salient to users can convey the context behind information search and assist users in identifying how information surfacing through different provenances could vary in credibility. The second study on the Twitter platform utilized a design intervention that offered users a direct window into parts of the provenance of information they were engaging with. Findings suggest that easy access to provenance can assist users in inferring specific judgments useful for credibility assessment. The third study on the TikTok platform revealed how users employ nuanced strategies unique to platform features to afford the socio-technical context of information. Findings suggest that when assessing credibility, users implicitly referenced the concept of provenance even without any direct provocation. My dissertation presents numerous contributions across data, empirical research, artifacts, theory, and methodology based on studies conducted at three distinct research sites. Two of these contributions are particularly noteworthy. Firstly, I introduce a framework that proposes how design features of various platforms can facilitate credibility assessment. This framework considers dimensions such as inauthenticity, unfavorable online associations, contentious behavior, lack of trust, and the potential negative consequences of sharing information. Secondly, my dissertation establishes information provenance as a platform-agnostic cue for signaling information credibility. I discuss the findings in light of modern-day media literacy principles to advocate that information provenance could serve as a platform-agnostic construct valuable for conveying the opacity of information to users, who, in turn, can employ that realization to assess information credibility.