Running a science channel or a fact-checking team
A continuing science channel needs formats suited to its audiences, reliable checking and publication processes, collaboration with trusted messengers and a way to assess reach and effects.
Platforms, formats and staffing
Unbiased Science began in 2020 when two scientists combined their separate efforts against COVID-19 misinformation, starting with Instagram Live sessions on personal accounts. Neither had formal communication training, and they developed their approach through experience and mistakes. Complementary population-health and biomedical expertise helped them explain a topic from different scientific perspectives, while academic, industry and teaching experience informed their audience choices.
By their 2023 account, the weekly podcast had passed one million downloads and provided detailed discussion of data. Instagram, Facebook, Threads and Substack carried infographics, short and long videos and long-form content, reaching different generations, ideologies and communities. The founders observed older, more politically right-leaning audiences for Facebook's long-form content and younger millennial parents for Instagram graphics and videos, and were expanding into TikTok for Generation Z. They adapted format and framing to education level and demographics, using direct, nontechnical language for people without scientific training who distrusted the medical establishment.
The channel expanded beyond COVID-19 to mammograms, vaping, wellness trends, genetic engineering, food ingredients and chemistry. Substack provided a lasting repository of podcast sources, while interns contributed research and infographic design. Its associated institute worked with public-health leaders, STEM students and community organisations through education, training and engagement to build scientific literacy and future communicators.
Technical and didactic material was supplemented with playful graphics, including a Schitt's Creek reference encouraging fruit and vegetable consumption regardless of labels, a Got Milk parody introducing dairy topics, and an ice-cream/shark-attack example distinguishing correlation from causation through their shared summer context.
Trusted messengers and campaign design
Where objections reflect systemic or cultural differences, trusted community organisations can carry information more effectively than an outsider announcing scientific authority. Scientists need to work with organisations, government, industry, communities and media, giving trusted partners information in a form they can convey. Message testing, where time and resources permit, can help predict impact.
EMA's first content-creator campaign, #HealthNotHype, ran for about a month from 4 November 2025 to promote safe and responsible GLP-1 medicine use. Creators from Finland, Germany, Italy, Latvia, Portugal, Slovenia and Spain were selected for geographical balance, healthcare-communication credibility and alignment with evidence, respect, transparency and independence; most were healthcare or nutrition professionals. They published audiovisual material and engaged followers through polls and quizzes, primarily on Instagram because of its volume of GLP-1 discussion.
Campaign reach and estimated effects
Unbiased Science's flu-vaccine campaign used surveys, focus groups and key-informant interviews with the public and healthcare providers to identify information gaps and hesitant audiences. Two groups emerged with different concerns: Black Americans of any age or gender described historically grounded mistrust and fears about chemicals, needles and injections, while young White men in rural Southern and Midwestern areas emphasised autonomy and government distrust. The team developed and tested two message approaches, then produced podcasts, infographics, reels and a live paediatrician question-and-answer session.
The organisers reported reaching 1.3 million people, including more than 300,000 clinicians who could pass information to the public. They estimated that 66,000–215,000 of those reached had not planned vaccination that season, averaging about 140,000; 15% changed their minds, 12% were vaccinated and 8% attributed the decision to the campaign. Their resulting estimate was 5,400–17,500 vaccinations in response to the campaign, averaging about 11,200, described aloud as more than 11,000 people changing their minds. At the 2023 presentation, they were preparing the results for peer-reviewed publication.
A menstrual-health campaign addressed medical questions about bowel movements, blood colour and overnight tampon use alongside fears about ingredients such as titanium dioxide, explaining products and their components. Podcasts, infographics and reels included a playful wombat image referring to its cube-shaped droppings. The two menstrual-myth podcast parts had 18,300 and 19,400 downloads. Reported Instagram figures were 74,000 impressions and 3,800 reactions for one set, and 34,000 impressions and 1,000 reactions for another.
A fact-checking team's workflow
Argentina's Science Anti-Fake News team began during the COVID-19 pandemic with sixteen researchers and doctoral students, becoming the first COVID-19 misinformation project endorsed by CONICET. It added twelve researchers in social communication, political research and anthropology from across the country and had completed more than 300 checks by its 2022 account.
Questions and claims arriving through social media were checked against scientific evidence and scientific-society consensus, then verified by several team members in a process resembling peer review. Peer-reviewed articles were the main sources; preprints were treated cautiously, explicitly labelled unreviewed and analysed closely where a media claim depended on them. The team monitored PubMed, bioRxiv, medRxiv and SSRN and consulted WHO, CDC, EMA and local health-authority guidance. Members wrote a sourced explanation, revised it into accessible nontechnical language and published it on the government ConfiAR platform and Instagram, Facebook, Twitter and YouTube. In 2022 it reported more than 30,000 Instagram, 6,000 Facebook and 24,000 Twitter followers.
Checks were chosen through audience interaction and news coverage. The team followed Cortassa's ethnographic contextual model, using dialogue, audiences' interests and existing knowledge, rather than the deficit model's assumption of scientific illiteracy.
Its three published examples reported the following reach:
| Claim checked | Date | Profiles reached | Likes and reactions |
|---|---|---|---|
| COVID-19 vaccines were developed too quickly to be safe and skipped preclinical stages | October 2020 | More than 30,000 | More than 2,000 |
| Alcohol must be avoided for 42 days after Sputnik V vaccination | December 2020 | More than 60,000 | More than 5,000 |
| A large proportion of COVID-19 vaccinations caused severe side effects | January 2021 | More than 50,000 | More than 3,000 |
The October misinformation also claimed vaccines changed DNA and was spread by some healthcare workers and false experts through radio and television. The December check reported that moderate alcohol consumption would not affect vaccine efficacy, while abuse could suppress immune responses; a televised correction gained 18,427 YouTube views. The team hypothesised that approaching holidays, unpublished phase III results and negative vaccine coverage explained its popularity. The January check countered viral WhatsApp claims of more than 20% severe effects using trial results and Argentine health-ministry reports showing no more than 1% at that point.
The team could not measure effects on television or radio audiences, and overlapping followers across platforms could overestimate social reach.
Sustaining and extending a channel
In 2023, Unbiased Science sought a hub connecting scientific, clinical, writing, media and regulatory experts with community leaders and legislators. Its founders wanted to make the work their main occupation because running it alongside full-time jobs required substantial time, energy and expertise.
In examples documented in 2021, David Attenborough briefly joined Instagram to reach younger audiences and set a record for the fastest million followers, while Raven Baxter combined science-themed music and rap to encourage public engagement.