top of page

Fake News: When in doubt, don't share, is a research project aimed at investigating the phenomenon of fake news durring the first semester of the COVID-19 pandemic, identifying the strategies used in mass manipulation and trends in the visual composition of the content.

Como as imagens enganam


This reaserch was born with a question that emerged from the Cambridge Analytica scandal in 2018:

Have I been manipulated?

This question started a process to discover what fake news is, how it spreads, but most importantly why we believe it, and what is the role of the non-verbal language—in this case visual language—in the process of manipulation. The objective was to discern if different types of imagetic fake news were more misleading than others.

There has always been disinformation, but Fake News is a unique phenomenon born out of the Age of Information. By pressing a button on a screen, news is instantly shared, reaching large audiences—that which the printing press can no longer reach.

The term was popularized in the US presidential elections in 2016. At the time, the Republican candidate and former president, Donald Trump, discredited news criticizing his governmental project, planting doubt in the minds of the population, to the point of creating a parallel reality.

“The use of a foreign expression for something that has a perfectly known name in our language (fake news), already indicates something beyond the normal meaning of the word.”
(PINHEIRO, 2019, p.88)

O que são

There are many components involved in the spreading og fake news. There's belief that traditional press in no longer reliable, which motivates the reader to seek alternative means of information; there are social network algorithms that help boost inaccurate content, there are  false content creators hired to develop inflammatory speeches and ideas by third parties, there's the use of artificial ads with agendas impulsed by bots, there's the user, who never learned to filter all of the information. And it all happens in the no-mans-land of the digital world.

“The movement of fake users has the power to guide the debate. The appearance that a topic is being debated makes gives the impression that it's in fact relevant.”
(CARVALHO, 2020)

Por que acreditamo

Throughout the research process, 5 factors were mapped, which help justify the adherence of manipulated content:

1. Familiarty;
2. Authority;
3. Psychology;
4. Narrative;
5. Presentation.

Similar content is repeated, generating a sense of familiarity, therefore giving credibility to the information.

Who shares content gives a sense of credibility, and the use of fake  experts or witnesses within the content.

The human mind helps to distort our preception of reality through heuristics and cognitive dissonance, among other psycological concepts.

By employing narratives that incite strong emotions (disgust, fear, or anger), fake news is shared immediately.

The quality of the visual composition of fake news and the arrangement of specific signs can deceive the eye and give credibility to the content.

Imagetic fake news were collected durring the project's research period and then analyzed. They can be classified into two trends according to visual language: Social Midia Aspects and Graphic Manipulation.

Analise visual
Análise visual ii

Based on this first classification, the analysis focused on fake news with graphic manipulation, generating questions such as: what makes an imagetic Fake News deceive its viewers? Is it possible that there are fake news with a greater potential for manipulation than others?

To answer these questions, a basic protocol was designed for the qualitative analysis, evaluating technical aspects and the quality of the composition as a whole.

Social Midia Aspects

Utilize predefined social media content creation resources and tools.

graphic Manipulation

Generated in an image creation and manipulation software.

With the protocol in place, it was possible to trace the a quality of the composition (low, medium, high), and its potential for harm (higher, medium, lower), through the analysis of 6 imagetic Fake News.

A new question arose: is there a relationship between the quality of a composition and its potential to deceive? The following 6 compositions were analyzed and organized into a graph that compares the quality of the composition with the potential for manipulation.


Throughout the course of this research, it was possible to map some crucial elements in the dissemination of fake news and how it affects our vulnerable human minds. When analyzing the visual language of the collected imagetic Fake News, more questions arose than answers.

Why is a low quality fake news composition  believable to certain audiences? Three theories were drawn to explain this: the user has a low visual literacy (has some difficulty in reading and interpreting the signs in the image); they have desensitized themselves to low-quality content (which is part of the internet's digital and participatory culture); the user really believes in that content and the shared image serves to reinforce the belief, ideology, etc. These questions, however, will only be answered at a later date.

However, when we consider high quality fake news, the ethics of designers and advertisers are called into question. As seen, imagetic misinformation with high quality compositions have the potential for greater feigned credibility, which betrays the trained eye of a visual comunicator.

Although the project has been completed, there is a need to continue studies on this subject, and to look deeper into the false visual content and who produces it.

As a possible solution to this very complex issue of misinformation, the urgency of rethinking the educational system and adding digital literacy to the basic education curriculum in Brazil is imperative. The more knowledge people have about how to browse the internet safely and consciously, the better their chances will be in fight against misinformation.

This project was presented at the virtual meeting of Design Activist 2020 can be seen here.
Download the full report here.

bottom of page