Researchers: Video may be more believable than text, but not necessarily more persuasive

Justin Hendrix / Nov 19, 2021

There is widespread concern about how emerging media may be used to manipulate political opinion. To better understand the impact of certain media modalities on political persuasion, researchers at MIT set out to interrogate the "widely held intuition that video is more believable and compelling than text."

In a paper published in the Proceedings of the National Academy of Sciences titled "The (minimal) persuasive advantage of political video over text," the researchers hypothesized that the media modality of video, as opposed to text, might influence whether people believe that "the information being conveyed is "genuine"-- in other words, testing the old adage "seeing is believing." And second, they set out to explore whether the information contained in video is more or less persuasive than text.

To get at these questions, the researchers conducted two surveys with a total of more than 7,500 respondents in the Spring of 2021. In the first survey, respondents were presented with videos and transcripts from a corpus of political ads related to issues ranging from climate change, healthcare and the minimum wage. In the second, respondents were presented with content about the COVID-19 pandemic, "each of which was widely viewed on YouTube during the peak of the pandemic in the United States." Some of the clips in the second survey were overtly political- such as "an MSNBC clip of former President Trump discussing testing rates," while some were not.

Importantly, the study compares the video to text artifacts that were presented as "a detailed transcript containing an exact replication of the audio output as well as a comprehensive description of key visual cues," and the researchers note that typical news articles or opinion pieces are normally presented quite differently.

The researchers find that "the effects of video versus text are remarkably similar across" the two surveys, with a respondents regarding information presented in video more authentic than that presented in text. Thus, "video seems modestly but meaningfully more believable than text."

But, as to the second hypothesis, it does not appear that video is notably more persuasive or engaging than text. "In sum, although video may be more believable than text, this enhanced credibility does not seem to be accompanied by a commensurate increase in persuasion or personal engagement."

The authors do note that the experimental conditions may not translate to the real world, particularly when a media artifact is presented in social media feeds:

In particular, it is possible that video is more attention grabbing than text, such that people scrolling on social media are more likely to attend to and therefore be exposed to video versus text. As a result, even if video has only a limited persuasive advantage over text within a controlled, forced-choice setting, it could still exert an outsized effect on attitudes and behavior in an environment where it receives disproportionate attention.

The researchers suggest future work should look more at these effects with regard to manipulated media:

In particular, although we observe clear effects of video on belief in unmanipulated content, future work should explore whether our results replicate in the context of misinformation, including deepfakes as well as so called “shallow” or “cheap” fakes that use simple editing tricks to craft deceptive footage.

Indeed, while some coverage of the study has suggested it indicates manipulated media-- such as deepfakes-- should be less of a concern because it may be no less persuasive than text, this finding is not present in the study per se. There are any number of critical contexts were a viewer's false conclusion that a video represents an authentic version of events or information may be impactful even if it does not have a persuasive effect on a viewer's views on a political topic.

But one of the study's authors sees this research as a precursor to understanding the potential impact of manipulated video.

"We got into this research through an interest in deepfakes, but we regarded these questions as more fundamental," said David Rand, a Professor of Management Science and Brain and Cognitive Sciences at MIT, and one of the authors of the paper. He says the paper "doesn't say deepfakes are not a problem, everybody go home." But it may indicate their potential as tools of political persuasion are less than some have feared.

"I'm less worried about deepfakes than before I did this research, but I'm not not worried," said Rand.

The research was funded with a gift from Jigsaw, a division of Google.


Justin Hendrix
Justin Hendrix is CEO and Editor of Tech Policy Press, a new nonprofit media venture concerned with the intersection of technology and democracy. Previously, he was Executive Director of NYC Media Lab. He spent over a decade at The Economist in roles including Vice President, Business Development & ...