The Visual Media Lab aims to bring together faculty and graduate students to examine the (1) attributes, (2) prevalence, and (3) impact of visuals in journalism and social media in the digital age. The visual lab is led by Alex Scott, who specializes in the qualitative examination of visual content and photojournalism; Sang Jung Kim, who specializes in the computational examination of visual content; and Bingbing Zhang, who specializes in studying the effects of visual content on audience perceptions, attitudes, and behaviors. Sachith Seneviratne, a research fellow at the University of Melbourne, is an affiliate faculty of our lab. The initial project within the visual media lab centers on understanding the occurrence and consequences of visual misinformation, particularly with a focus on deepfakes. Below we elaborate on the 1) initial project’s scope, 2) detailed project descriptions, and 3) project implications.

Scope

Visual misinformation, such as deepfakes, has the potential to greatly influence people's beliefs, attitudes, and behaviors due to its striking characteristics. Our research employs mixed methods approaches to investigate visual misinformation and its impact. We aim to understand how both the public and the news media perceive deepfakes, analyze their distinctive attributes, and identify the factors contributing to the widespread and persuasive nature of visual misinformation. Additionally, we plan to develop debunking and inoculation strategies to counter deepfakes and explore their effectiveness.

Aim1: Understanding the current scope of journalists and public understanding on visual misinformation (e.g., deepfakes) using computational analysis.

Aim 2: Qualitative examination of visual misinformation.

Aim 3: Testing the effects of visual misinformation & effects of strategies to deal with visual misinformation

Implications

Implication 1: Building a theoretical framework to examine the public discourse, features, and effects of visual misinformation.

Implication 2: We aim to assess the current visual media ecosystem by detecting and analyzing instances of visual misinformation in both journalism and social media.

Implication 3: Inform the strategies to deal with visual misinformation for public policy makers.