Researchers break down the digital habits of science influencers

A new study exploring how scientists share knowledge on social media reveals that different digital platforms and scientific topics require distinct communication styles to reach wide audiences. The research highlights that while snappy, objective videos perform best on TikTok, detailed and slightly ironic explanations gain more traction on YouTube. These findings were recently published in the journal Computers in Human Behavior.

In our highly digital society, the way scientific knowledge reaches the general public relies heavily on social media networks. Online platforms allow experts to speak directly to citizens without filtering their messages through traditional news media. This direct line of communication helps combat widespread misinformation and builds a more informed public.

Yet, the sheer volume of content available online makes it difficult for users to identify reliable sources. Many science communicators post accurate, well-researched information that fails to attract viewers. Researchers wanted to understand the specific elements that help certain scientific posts go viral while others remain unseen.

Knowing what drives audience engagement helps content creators design better educational material. Lead researcher Montserrat Aiger of Zaragoza University in Spain and her team set out to identify the common traits of successful science videos. They focused their investigation on the precise characteristics that determine digital popularity across different scientific fields.

Social media platforms offer a unique opportunity for academics to build trust with everyday people. Before the internet, scientific discoveries primarily lived in dense academic journals or specialized magazines. Now, a biologist or physicist can record a quick video on a smartphone and reach millions of viewers in hours. This modern shift relies on communicators acting as relatable figures who can break down complicated concepts.

The research team examined the performance of prominent science influencers on Instagram, TikTok, and YouTube. They selected sixty total accounts representing five distinct fields of knowledge. These fields included arts and humanities, health sciences, experimental sciences, social sciences, and engineering.

To gather their data, the researchers pulled the twenty most popular videos from each selected profile. This process generated a total sample of one thousand two hundred videos posted between early 2022 and late 2023. The team then used analytical software to measure engagement metrics like total comments, user reactions, and overall view counts. The researchers established a specific time window to capture a broad sample of digital activity while comparing the different applications fairly.

After downloading the numerical data, the team turned their attention to the spoken words. They utilized voice recognition software to transcribe the audio tracks of all the videos. Once the transcripts were complete, the researchers fed the text into a linguistic analysis program.

This program evaluated multiple aspects of the language used by the influencers. It measured polarity, which determines if a statement leans heavily positive or heavily negative. It also checked for subjectivity, dividing statements into either factual descriptions or personal opinions. Finally, the software looked for instances of irony and emotional agreement within the sentences.

The investigators discovered distinct behavioral patterns across the three social media networks. TikTok consistently generated the highest levels of audience interaction, including likes and comments. Videos on this app leaned toward highly objective language mixed with intense emotional extremes. The short format seemed to reward fast, attention-grabbing delivery over nuanced debate.

Instagram appealed to slightly different user habits. The app rewarded visually appealing content coupled with positive emotional framing. Catchy captions filled with popular hashtags and optimistic language performed exceptionally well on this network.

YouTube produced entirely different engagement patterns compared to the other two platforms. Although YouTube videos generated fewer total likes and comments per post, they provided a space for deeper, longer explanations. Creators on YouTube successfully used higher levels of subjectivity and irony to keep their audiences entertained during extended discussions.

The research also demonstrated that audience expectations change depending on the branch of science being discussed. Viewers seeking out information about physics, mathematics, and engineering tend to reward objective presentations. Creators in these fields generated massive numbers of likes by clearly explaining concepts without relying on heavy humor or subjective opinions.

This pattern flipped entirely within the social sciences. When creators discussed topics like psychology and sociology, their audiences wanted to chime in with personal thoughts. Videos in this category generated relatively lower like counts but experienced an explosion of comments. The communicators who leaned into subjective opinions and ironic commentary successfully stimulated debate among their viewers.

Health science accounts found success by maintaining an objective tone while remaining relatable. Their highly engaged audiences responded well to factual medical information that lacked heavy irony. Experimental science creators performed best when they mixed positive emotions with moderate portions of humor and subjective opinion.

Arts and humanities profiles tended to score closer to neutral on the emotional polarity scale. Because these topics often invite deep reflection rather than quick reactions, their overall engagement metrics lagged slightly behind the hard sciences. These creators still managed to build dedicated followings by using irony and highly subjective messaging.

Based on the data, the researchers suggested several guidelines for aspiring science communicators. Those hoping to grow an audience on TikTok should focus on short, visually dynamic videos that reach their point quickly. Emotional appeals have less impact on this application, so creators should focus on prompting immediate interactions from their viewers.

For Instagram, communicators should pair their educational facts with an inspirational tone. The platform rewards personal storytelling, which helps academics build a sense of proximity with the public. YouTube creators should worry less about gathering quick likes and focus instead on crafting structured, educational lessons. Using humor and emotional diversity can help keep long-form content engaging.

While the results offer clear guidelines for online educators, the researchers noted several limitations to their work. The study focused heavily on the most popular posts from established influencers. By excluding less successful videos from smaller creators, the study creates a certain level of survival bias. What works for a massive influencer might not produce the same results for a new account.

The design of this analysis also means it cannot prove that specific emotional tones directly cause videos to go viral. The invisible algorithms guiding these platforms frequently prioritize content based on hidden corporate formulas. Social media companies design their software to keep users watching, meaning audiences might not be freely choosing the content they consume.

The reliance on automated text analysis also carries potential biases. Artificial intelligence tools sometimes struggle to interpret cultural slang or subtle sarcasm accurately. Because the videos tested featured variations in language and regional dialects, the software might have missed some of the more nuanced human emotions.

The metrics themselves provide only a superficial measure of educational success. A high number of likes shows that a video captured attention, but it does not reveal what the viewer actually absorbed. Tapping a screen requires very little mental effort. True learning demands a level of deep engagement that simple social media metrics cannot easily measure.

The study, “Science communication in social Media: Analysis of success on TikTok, Instagram, and YouTube across scientific disciplines,” was authored by Montserrat Aiger, Carmen Elboj, Raquel Lozano-Blasco, and Marian Acero-Ferrero.