Twitter is an increasingly powerful tool for information. In April, a hoax tweet from someone who hacked the Associated Press Twitter account caused the stock market to momentarily crash. Traders were swayed by a single false tweet about a terrorist attack on the White House. The incident begs the question of how we evaluate information in this social media era. A new study suggests, however, that we are more immune to remembering false information on Twitter than we may think.
“We don’t need to think of social media as dumbing us down to the point that we believe everything we read on Twitter,” says Susan Ravizza, a cognitive neuroscientist at Michigan State University. “People are treating information presented on social media with well-deserved skepticism.” Those are the results of a new study, just published in Psychonomic Bulletin & Review and widely covered in the news, that looked at how Twitter affects false memory formation.
In the study, the researchers – led by Kimberly Fenn, also of Michigan State, and Ravizza – presented to undergraduate students a series of 50 images that told the story of a man robbing a car. After the participants performed an unrelated memory test to distance them from the original story, they all viewed an information feed of 40 text lines narrating the events in the earlier-seen images.
Most of the information in the feed was accurate to the original images except six details that were contradictory. For example, one line of text read, “The car had a Harvard sticker in the back window,” when the car in the images actually had a John’s Hopkins sticker.
The researchers assigned the participants into one of three groups for viewing the information feed: two made to look like a Twitter feed (labeled “Tweet ticker” with a light blue background and Twitter logo) and one control feed (labeled “Photo recap with a red background”). One of the Twitter feeds used more formal sentences and language like the control feed, while the other was written informally like a more traditional Twitter feed. All the participants then took a test about the information they saw in the images, to see if any of the false information planted in the feeds took hold.
Overall, the researchers found that the young participants who read the Twitter feeds were less likely to form false memories about misleading information than those who read the control feed. The formality or informality of the language in the Twitter feeds made no difference as to how confident the participants felt about the correct, or incorrect, information.
Ravizza talked with CNS about these findings and their implications for how we use social media and form memories.
CNS: Why did the study use Twitter instead of other social media sites, such as Facebook?
Ravizza: People who gather information through Internet news outlets, television, or newspapers do not know the reporters who are writing about events. We chose Twitter because it is a less personal form of social media than, for example, Facebook. Facebook friends are more likely to be personally known than Tweeters. Twitter was the most similar to gathering information through non-social media sources.
CNS: Your study was with undergraduate college students, who presumably have grown up with social media as a source of news and information. Does your research shed any light for older Twitter/social media users?
Ravizza: We have asked ourselves this as well and would like to run an experiment with older adults who have not grown up with social media. Older adults generally show higher rates of false memory because they are more likely to forget the source of the information. It is possible that they would not show an effect of Twitter on false memory because they have forgotten where they acquired the information. This would have to be carefully controlled so that we could see effects of social media apart from general aging effects on false memory.
CNS: Why do you think the Twitter format created less confidence in false information?
Ravizza: Marshall McLuhan famously said, “The medium is the message” and young adults may take this into account when reading information presented through Twitter. The Twitter format may be viewed as less trustworthy and more inaccurate compared to other sources of information.
CNS: How does this study fit in with the body of work on false memory formation?
Ravizza: Our work shows that false memory formation for social media is subject to the same principles as other types of sources. Sources that are viewed as untrustworthy or having low credibility are associated with lower rates of false memory formation.
CNS: Were you surprised by any of the results?
Ravizza: We were surprised by the results because we had been predicting the opposite effect! We thought that the informal language used in Twitter might engage more interest in its content. If so, false information might be more deeply integrated into people’s memory of an event and so rates of false memory would be higher. In fact, the formality of the language had no effect on false memory formation. We underestimated the skepticism directed at information presented on Twitter.
CNS: How did you become interested in the link between social media and false memory formation?
Ravizza: I enjoy posting on my personal Facebook account and I became interested in using social media as a tool to engage the general public in cognitive neuroscience research. My lab has a Facebook page, a Twitter account, and a YouTube channel! I was curious about potential advantages and pitfalls of using these tools in the science domain and included it as one of my aims in my National Science Foundation grant proposal.
CNS: What’s next for the work?
Ravizza: We would like to understand why Twitter is treated with skepticism. Is it the anonymity of the posters or their presumed lack of expertise or familiarity with the information they are discussing? We would also like to observe whether this is a generational effect or present at all ages.
–Lisa M.P. Munoz
The paper “The effect of Twitter exposure on false memory formation” by Kimberly M. Fenn, Nicholas R. Griffin, Mitchell G. Uitvlugt, and Susan M. Ravizza, was published in Psychonomic Bulletin & Review online on May 14, 2014.