Communicating Science and the Problem With “Context Collapse”
Sunday, October 22nd, 2023
This week’s core story is about: Context collapse.
KNEAD TO KNOW
A new study by researchers at the University of Pennsylvania found patients with long COVID can experience reduced levels of serotonin, offering a possible explanation for symptoms like brain fog, fatigue, and memory loss. The study found the serotonin depletion may be triggered by remnants of the coronavirus lingering in the gut, shedding new light on the mechanisms that may cause long-term neurological symptoms in COVID patients.
New data from the FBI’s annual crime report shows anti-LGBTQ hate crimes increased sharply in 2022, jumping 19% compared to 2021 estimates. The FBI found the number of anti-Jewish hate crimes also jumped 36% compared to 2021, though the majority of hate crimes targeted Black Americans (52%).
A new report by the National Academies of Science, Engineering, and Medicine lays out a comprehensive roadmap for the U.S. to transition away from fossil fuels and achieve its net-zero carbon emissions goal by 2050. The sweeping report focuses on societal objectives and technology, and includes more than 80 recommendations aimed at both the public and private sectors. The report follows a 2021 report by the institution that provided a “technical and federal policy blueprint” for the next decade, and helped inform the climate policies in the 2021 Infrastructure Investment and Jobs Act and 2022’s CHIPS and Science Act and Inflation Reduction Act.
A new report by Similarweb found one year after Elon Musk’s acquisition of Twitter (now X), traffic and monthly active users on the platform are down significantly, though traffic to Musk’s personal account is up. The report found global web traffic to Twitter.com in September was down 14% year-over-year, while combined monthly active users on its Android and iOS apps was down 17.8% year-over-year. Web traffic to the ads.twitter.com portal for advertisers was also down 16.5%. Despite that, traffic to Musk’s personal profile was up 96% year-over-year.
Communicating Science and the Problem With “Context Collapse”
New research investigates what happens when non-experts share science they don’t understand in an effort to seem right.
If you were paying attention to the news during the COVID pandemic (I’m assuming you all were), then you’ll remember the days of scientists and government officials scrambling for solutions as we faced one of the greatest threats humanity has seen in some time.
If you were paying attention to the news and also on social media, then you’ll also remember the countless right-wing influencers and media folks peddling things like Ivermectin and colloidal silver as bogus miracle cures.
One of the problems was that the people peddling bogus solutions like Ivermectin seemed real. While many of us familiar with science and research ignored the hype about a horse dewormer, many who didn’t know any better trusted the journalists and influencers promising solutions.
Said journalists and influencers seemed so certain.
They presented “studies” purportedly showing they were right and framed the findings in a way that fit their narrative. They pulled quotes that seemed to back up what they were saying and added screenshots of official-looking abstracts and title pages. And they did it all while throwing out terms like “peer-reviewed” and “clinical trial” to give a veneer of seriousness and consensus.
They engaged in what’s known as “selective reporting,” in which a person cherry-picks evidence to support their claims and manufacture the impression that their position is right.
New research investigates the implications of selective reporting in the age of social media and what happens when people who aren’t equipped to understand certain information try to reason about it.
What the research says:
A new study by researchers at the University of Washington examined how “science communicators,” here, a broad category of people who post about science on social media, including everyone from nefarious conspiracy theorists to seasoned journalists, can transform scientific evidence into competing narratives and manufacture the appearance of consensus. The team analyzed a dataset of 5 million tweets using online discourse about the effectiveness of masking during the COVID pandemic as a case study.
The study found science communicators, one side comprised of people from mostly mainstream academic and journalistic institutions and the other largely a coalition of conspiracy theorists and content creators, transformed the online perception of scientific evidence by selectively amplifying certain published work while diminishing others to create bodies of “evidence” that either supported or opposed masks, depending on the communicators’ view.
Examining published studies cited by both pro- and anti-mask science communicators, the researchers found they often cite the same papers, but in different ways.
Pro-mask communicators tended to quote or paraphrase top-level findings from a study and rarely used negative citation (i.e., citations going beyond discussing a study’s limitations to actively recommend readers disregard the work), while anti-mask communicators instead quoted a selective sentence from a study and used negative citation frequently.
Anti-mask communicators also quoted findings out of context, deceptively implying the opposite of a study’s conclusion and the author’s publicly stated opinions about the work.
By selectively quoting papers and/or deceptively implying certain conclusions, anti-mask science communicators added an appearance of legitimacy to their arguments, even when the authors they cited publicly supported the effectiveness of masking.
Why it matters:
The study provides context about a phenomenon called “context collapse,” where one group (non-experts) encounters information meant for a different group (experts) and therefore lacks the ability and knowledge to correctly reason about the findings.
The group of non-experts who encounter research in isolation often aren’t able to accurately assess scientific evidence in light of other work or broader domain knowledge, and therefore might not be reliable sources of information.
The study also has implications regarding the social construct of consensus by non-experts, which can cause anxiety among scientists who worry the public may come to view mainstream scientific consensus as not objective and therefore potentially untrustworthy.
The researchers are calling on scientists and scientific publishers to take certain steps, like allowing experts to amend contextual information to papers that are at risk of being misused. They also believe scientists, reviewers, and editors should be careful to craft their papers and abstracts in ways that aren’t easily misquoted.
The findings also underscore one of the more frightening aspects of the age we live in, one in which forming a factually supported consensus on important issues seems difficult when so many people on social media are willing to confidently share their wrong opinions (often to wide audiences of similarly incorrect people).
As a society, we must be able to form a consensus about vital issues like public health, climate change, and politics, among others, and move forward informed by that shared understanding.
Failure to do so looks a lot like what we saw during the COVID pandemic, when reality became distorted by false cures and outright lies. It cost lives during the pandemic; we’ll see what price we pay next time.
Supreme Court Justices Are More Likely to Interrupt Women Attorneys During Oral Arguments
A recent study by researchers at the University of Massachusetts Amherst found U.S. Supreme Court Justices during oral arguments are more likely to interrupt attorneys who are women than they are to interrupt attorneys who are men, underscoring the effects of gender on appellate advocacy. The study was posted on the preprint database SocArXiv. Reminder: preprint studies have yet to be peer reviewed and accepted for publication by an academic journal.
The researchers examined four decades of Supreme Court transcripts, finding women attorneys during oral arguments were interrupted an average of nearly one additional time per 1,000 spoken words when compared to men, with justices who are men more likely to interrupt women attorneys compared to justices who are women.
The study also found that while most justices included in the study were more likely to interrupt women than men, the effect was especially pronounced among conservative-leaning justices.
Los Angeles, CA; Getty
Cities in Blue States Are Safer From Gun Violence than Cities in Red States
A new report by the Center for American Progress found cities in blue states (based on how the state voted in the 2020 election) experienced an average gun homicide rate of 7.2 per 100,000 residents from 2015 to 2022, while cities in red states experienced an average rate of 11.1 per 100,000 (53% higher than blue-state cities).
Examining the 300 most populous U.S. cities, from 2018 to the peak of U.S. gun violence in 2021, the report found red-state cities saw their rates of gun homicides increase by 27% more than blue-state cities. Red-state cities also experienced 27% more accidental shootings in 2022 than blue-state cities.
The report also found gun violence is decreasing faster in blue-state cities (down 14.9% year-to-date) than red-state cities (down 3.7% year-to-date).
Who Exactly Lives in the American Midwest?
A new survey by researchers at Emerson College Polling and the Middle West Review asked residents of 22 states whether they consider themselves to be Midwesterners, embarking on what the authors describe as “the largest-ever study on Midwestern boundaries and identity.” The survey included the 12 states designated as Midwest by the U.S. Census Bureau and their surrounding states.
The survey found majorities of residents in Iowa (97%), Minnesota (97%), Missouri (95%), Illinois (94%), North Dakota (94%), Wisconsin (94%), Nebraska (93%), South Dakota (92%), Indiana (92%), Kansas (91%), Michigan (85%), Ohio (78%), Oklahoma (66%), and Wyoming (54%) consider themselves to “live in the Midwest.” Majorities in those states also view themselves as “Midwesterners,” though to a lesser extent.
“There are doubters out there who believe that the Midwest doesn't exist, or it's not a place where people have a strong sense of regional identity,” lead author Jon Lauck told St. Louis Public Radio. “And this study clearly rebuts both of those claims.
4.8 hours - The amount of time per day the average U.S. teenager spends on social media, according to a new survey by Gallup. The survey found girls spend nearly an hour more per day on social media than boys (5.3 hours vs. 4.4 hours, respectively).
8 billion years - The age of the Fast Radio Burst (FRB) detected by astronomers in Australia in June 2022, according to a new study by researchers at Macquarie University. The team found the source of the rapid burst of radio waves to be a group of two or three galaxies that are merging. It’s the most ancient and distant FRB located to date and could be used to “weigh” the universe.
48% - The percentage of U.S. Gen Zers who say they’ll do at least some of their holiday shopping on social media, according to a new survey by Shopify and Gallup. The survey also found 37% of Gen Zers say they plan to spend more money on gifts this season than last year, nearly double the average across all other age groups.
Short Video. Meet the man who designed history’s most famous tongue. (4 min)
Fun Video. What actually happens if you don’t put your phone on airplane mode? (5 min)
Good Read. Learn about Canadian inventor George Cove, one of the world’s first solar entrepreneurs whose 1909 kidnapping might have changed energy history. (1,019 words; 5 min)
Neat List. Check out 10 fascinating glimpses of the microscopic world.
SHARE THE BAGEL
Like what you’ve read so far? Don’t be selfish, share The Bagel! You’ll sound super smart and cool recommending The Bagel to your friends, family, coworkers, significant others, insignificant others, and everyone in-between!
You’ll also get a promo code for a FREE BAGEL at any Einstein Bros Bagels, Bruegger’s Bagels, or Noah’s Bagels location if you can get five people to subscribe using your unique referral link below!
Share your link: https://blog.readthebagel.com/subscribe?ref=PLACEHOLDER
You have: 0 referrals.
You need: 4 until a FREE BAGEL!
Heads up: The promo code is actually a $5 gift card that can be used at any of the three bagel shops. So, depending on what you order, you may be able to get TWO FREE BAGELS with this one gift card!
That's right. Potentially TWO FREE BAGELS.
Double up! Save it for later! Surprise a friend at work! Leave it in the car and forget about it until the smell gets you! The choice is yours. You've earned it.
Whatever you do, just remember: If you have them slice it like a loaf of bread like those monsters in St. Louis, don't tell them we sent you.
CONNECT WITH ME
Feedback, comments, or suggestions? Reply directly to this email (I try to respond to everyone) or connect with me on Threads.
Did a super smart and cool person forward this email to you? Head to our website to get it every Sunday.
Written by Ryan Wittler