From ScienceMag:

The U.S. National Science Foundation (NSF) has chosen a record number of students this year to receive its prestigious graduate fellowship, rebounding from last year’s unusually small cohort. The size of this year’s class, announced today, together with a more traditional distribution across fields, could ease fears that NSF, under pressure from President Donald Trump’s administration, had decided to shrink and alter the nature of a program that has supported 50 future Nobel laureates since it began in 1952.

“My take is that the STEM community’s activism around last year’s cuts appears to have had significant positive impacts on this year’s class,” says Susan Brennan, a cognitive psychologist at Stony Brook University and former fellowship program officer.

The 2599 fellows in this year’s class surpass the previous record of 2554 in the 2023 cohort and is 42% larger than last year’s unusually small class of 1500. The 2025 class was announced in two stages, and the second cohort of 500 was dominated by artificial intelligence and quantum information science, top priorities for the White House.

This year’s class returns to a more familiar distribution, with engineering and biology once again leading the pack. But whereas in 2023 and ’24 those two disciplines each garnered about one-quarter of the awards, this year saw a shift toward engineering, with some 35% of the awards going to students in that area, followed by 19% in biology. At the same time, the life sciences was the discipline most represented among the honorable mentions, comprising 40% of the 1440 total. Among awardees, computing and information science saw a slight uptick, from 7% in 2023 and ’24 to 10%. And psychology took a hit, from 5% and 6% to 2% of the overall pie.

The NSF fellowship provides students with an annual stipend of $37,000 for 3 years and gives their institutions $16,000 annually to defray tuition and other educational costs. It’s also a portable scholarship, in contrast to the typical arrangement in which a graduate student’s support is tied to their institution, either from their adviser’s research grant or a graduate training program in a particular field. NSF says it received nearly 14,000 applications for this year’s class.

NSF has launched an initiative to identify high-tech companies willing to contribute to the support of future classes of graduate fellows. But an agency spokesperson says, “NSF plans to support [this year’s] fellows with available appropriated funds in [fiscal year] 2026.”

Read More

From ScienceMag:

“Have you considered running for elected office?” My friend’s question didn’t come out of nowhere. I was active in my community as a volunteer, especially in environmental and social justice causes, and I regularly met elected officials and advocated for issues I cared about. But the question still took me by surprise. As a tenured professor and dean, my academic identity was firmly established. Was politics even something that academics did?

By the usual measures, I was successful. I had good funding, a solid publication record, and I had been promoted to serve as dean of engineering at the liberal arts college where I work in New York state. I enjoyed my leadership role and my research. But I did have reason to think about moving in another direction.

My most cited paper was a nice article with some juicy math—3D vector calculus in non-Cartesian coordinates!—but the work had little relevance to everyday issues. That always bothered me.

So did academics’ reluctance to speak out about policy. I had noticed that even when scientific papers did have findings worth sharing with the public or government officials, they tended to bury phrases like “We recommend that policymakers do X” near the end. There was an unstated assumption that a scientist’s role is to inform policy, not help enact it. That stuff was done by other people.

And as a researcher who’s had his share of scientific disagreements with other researchers, I have been able to work with others whose viewpoints differ from mine—an approach that is needed in these times of intense political polarization.

• Ashok Ramasubramanian
• Templeton Institute at Union College

When I turned 50, I also started to ask myself uncomfortable questions about my own future, such as, “What can I do with the time that is left to me?” I wondered whether I would have regrets if I did not serve my community more directly. After fulfilling my dean duties, I only had so much time left in the day. I realized I could not be an active researcher and engage in public service. To make it work, I would have to give up research.

I had just completed work on a major federal grant. So, I began to think the time might be right to consider running for a town council position. When the idea was only a nascent possibility, I broached it with my boss, our college’s vice president of academic affairs, and was pleased to discover that she was supportive. Our institution encourages community service and outside-the-box thinking, and administrators are generally happy for faculty to branch out. The idea is to help model lifelong learning—a value we work to instill in our students.

After much thought, I decided to take the leap. I closed down my lab space and liquidated all my research assets, turning them over to more junior faculty members, and began to spend my nondean hours going door to door and talking with voters.

It was a new world, and I had to learn a lot of new things quickly. My experience dabbling in research fields outside my own was helpful as I tackled activities that were new to me like fundraising, campaign finance reporting, and social media outreach. But I also leaned heavily into my favorite philosophy: “Fake it till you make it.”

I was pleased to find that voters in my community appreciated my candidacy. Being a scientist and an academic helped me stand out as a unique and qualified individual. And after I won my election and was appointed to the town council in January, I have tried to use skills I gained as a scientist to help my community. For instance, my experience writing research proposals is helpful when applying for grants aimed at infrastructure maintenance and green space preservation. And as a researcher who’s had his share of scientific disagreements with other researchers, I have been able to work with others whose viewpoints differ from mine—an approach that is needed in these times of intense political polarization. The time commitment has also been manageable, as council meetings are held in the evenings after normal work hours.

I miss many aspects of research, especially spending quiet time in my lab and mentoring students. But the experiences of running for office and serving the public as an elected official have been equally rewarding and fulfilling. I am not sure what my political future holds, but for now I am having quite a bit of fun serving my community in an official capacity. I encourage other scientists to ask hard questions about new ways to put their skills to work, especially in the second half of life.

Read More

From ScienceMag:

My heart raced as I walked into the classroom, where 200 curious medical students were waiting for me. While the technician fitted my microphone, I gripped the podium and scanned the sea of expectant faces. After years of turning down opportunities to teach, I’d finally agreed to give it a go—and I was terrified. I dove in and felt myself whizzing through my slides, trying to get through the material before my nerves got the best of me. After a few minutes, a student raised her hand and asked me to slow down. I felt my face go red—had I messed up my first ever lecture?

I never imagined I would find myself in front of a classroom. As an introvert and a nonnative English speaker, I found the prospect daunting. I’d see colleagues face hundreds of students and give fluent, engaging lectures and think I could never match up to them. Instead, I decided I’d stick to research, where I felt comfortable running experiments, applying for grants, and mentoring individual students.

A few years into my postdoc, however, my mentor asked whether I could teach her class while she sat on a grant-review panel. Out of respect for her, I said I’d give it a try, though I was nervous. I spent hours preparing, listening to recordings of her past lectures and cramming her slides with extra information, worried I’d forget what to say.

So I was embarrassed when, just minutes into the lecture I’d so meticulously prepared for, the student told me she was having trouble keeping up. I paused, took a breath, and adjusted my pace. And to my surprise, the energy in the room shifted. Students leaned in and asked questions, and I began to feel more of a connection to what I was teaching. I was back on track. By the end of the class, nothing had gone terribly wrong and I was relieved to have fulfilled my obligation to my mentor.

Later, when I watched the recording of the class, I could see my teaching get better as the lecture went on, and I began to get excited by the prospect of improving further. With my mentor’s support, I decided to take on more classes.

And so a single lecture grew into a regular commitment and eventually a responsibility I embraced. It took time and practice to become a confident, engaging teacher, but student feedback and teaching evaluations helped. After a student told me my slides had too much text, for instance, I redesigned them to include more visuals and fewer words, and found that this change helped make discussions more interactive. Eventually, I was offered a position designing courses as well as teaching them, something I had never anticipated in my career path.

At first, I worried teaching would distract me from the relentless demands of maintaining a funded research lab. But I actually found it sharpened my focus and transformed how I communicate science to colleagues and funders. Preparing lectures required me to revisit fundamentals I hadn’t thought about in years, keep up-to-date with new science, and learn to clearly explain complex ideas. In the lab and at conferences, I slowed down and focused on explaining concepts and protocols clearly, resulting in better discussions and more collaboration. I even secured a major grant—proof that clarity and connection matter as much in funding proposals as in classrooms—and my teaching experience helped me gain an earlier than anticipated promotion to my next faculty appointment.

Looking back, saying “yes” to teaching was one of the most transformative decisions of my career. It didn’t just make me a better educator; it made me a better scientist. For anyone nervous about the prospect of teaching, I can only recommend giving it a go. It’s common to worry about language fluency, feeling exposed in a room full of brilliant minds, or being pulled away from research duties. But anyone who thoroughly understands their subject can become a better communicator with practice and by refining their approach over time. Sometimes the most fulfilling academic life is not the one we first imagined, but the one we build through both intentional choices and unexpected experiences.

Read More

From ScienceMag:

It is a truth universally acknowledged that any scientific conference, no matter how fascinating, will become a snoozefest—usually during the time slot just before lunch.

But during one such conference a few years ago, Stefano Mammola, an ecologist at the Italian National Research Council, serendipitously found a loophole to this iron law. After sitting through several dull talks, he started chatting with fellow attendee Victoria Stout, an environmental scientist at the University of Colorado Boulder who moonlighted as a stand-up comedian. They quickly realized they had both made the same observation: Whenever a speaker cracked a joke, the audience instantly became more alert and engaged, the speaker appeared more approachable, and the talk itself became more memorable. “When somebody uses humor in an effective way, I recall the information much better in the future,” Mammola explains.

The pair realized they had the makings of a research project on their hands: a comprehensive analysis of how scientists use humor as they relay their findings. “Scientists attend many conferences,” Mammola says. “Why not collect some data?” Over the next few years, the two researchers—along with a growing number of interested collaborators—attended 14 biology-related conferences, collecting data on the use of humor across 531 talks.

As the team reports today in the Proceedings of the Royal Society B, scientists take themselves very seriously indeed: Most presentations contained no jokes at all or just a few. When they did occur, jokes tended to cluster at the beginning and end of presentations, and the majority either fell flat or elicited only polite chuckles—although the authors noted a bump in successful jokes midway through talks. Men speakers told more jokes, and jokes from men and native English speakers tended to get more laughs.

Mammola spoke to Science about the findings and about the potential of humor to improve science communication. The interview has been edited for brevity and clarity.

Q: How did you decide what counted as a successful joke?

A: We really wanted to capture any deliberate attempt to make people laugh, whether it was orally delivered, a gesture, or visually depicted—a meme in a slide, for example. The last two categories were more obvious, while the first was a bit more subtle. But when I was sitting together with Victoria Stout and trying to score independently, we realized that we mostly agreed when somebody was attempting a joke.

We also assessed joke success, which was not easy to standardize. We used this three-level breakdown, where if nobody was laughing or just a few people, that’s low success. Medium success is more or less half the room, and high success is when more or less the whole room starts laughing.

Q: Which jokes were the most successful?

A: We didn’t find any pattern with respect to the type of joke. Of course, some types of jokes are more frequently used, but there was not a single type of joke that was more successful than others. Anything can make people laugh or not—it’s more the delivery and timing.

In general, jokes cluster at the beginning and end of presentations. As you start, you want to engage with the audience and connect with them, or maybe you’re a bit nervous, so you throw in a joke to break the ice. And then at the end, you relax a bit. Maybe you want to leave people with a good, lasting impression.

This pattern was quite ubiquitous among all groups, except students didn’t joke as much at the beginning of talks. This category is the one with less public speaking experience, so they may be more nervous. They had the same peak at the end, so as the presentation progressed, that nervousness probably went away, and they managed to catch up.

Q: You also saw a bump in successful jokes halfway through the presentation. Why do you think that is?

A: When you are speaking to an audience, you realize at some point that you’re starting to lose them. Their minds start to drift. It’s inevitable. And I think an experienced speaker, about halfway through a presentation, is able to throw in a very nice joke to re-engage the mind.

Q: What other trends did you observe?

A: Male speakers said more jokes on average, and jokes delivered by male and native English speakers tended to attract more medium- and high-intensity laughter. Are they really better at telling jokes, or is it that people are more willing to laugh? Joking is a risky activity, because we have this idea that scientists should be serious, and the ability to take risks is not equally distributed. It’s a powerful reminder that inequality in academia affects so many things. I think part of the solution is changing the status quo, discussing these issues, and exposing them.

Q: What do you want people to take away from this research?

A: One conclusion is just the importance of thinking about science communication. The information system in science is increasingly polluted. There are so many papers, so many conferences, so many talks, so much information. The ability to stand out from the crowd and effectively engage your audience is really important and something we need to actively think about.

Q: Of all the jokes you heard, do you have a particular favorite?

A: I cannot come up with a single joke, but what is most effective to me is when people use their bodies, when there’s something totally unexpected in the way the speaker moves. To me, these are the most successful. I also really got to dislike all the stereotypical jokes from speakers talking right after lunch. I guess it’s inevitable, but my data tells me it doesn’t work. You have to be creative.

Read More

From ScienceMag:

Nearly half of biomedical scientists worry preprints could spread shoddy research and misinformation, according to a new survey that could help explain why the life sciences have taken up the publishing practice more slowly than some other fields.

The survey is one of the largest to date to examine views of life sciences researchers on the practice of placing non–peer-reviewed manuscripts on public servers. The results, posted this week on the bioRxiv preprint server, also reveal that researchers on average do not believe publishing preprints enhances their career advancement. But many acknowledge benefits, such as spreading their findings more quickly than peer-review journals do and helping them find collaborators.

“This study makes a valuable contribution because it highlights the persistent tension between the benefits of rapid dissemination and the way research is evaluated,” says Jeremy Ng of University Hospital Tübingen, who studies health research methodology and was not involved in the new study. “Hiring, promotion, and funding decisions often still revolve around traditional journal publications.”

Biomedical preprints have become more common over the past decade and spiked during the COVID-19 pandemic. But previous studies have indicated larger shares of physicists and economists regularly post preprints than researchers in the life sciences. “We wanted to know what is stopping the [biomedical] community from adopting them to a larger extent,” says information scientist Chaoqun Ni of the University of Wisconsin–Madison, who led the new study.

The survey, completed by nearly 1800 biomedical researchers in the United States and Canada in early 2025, reveals substantial variations in the use of preprinting. Two-thirds of respondents read at least one preprint during the previous 2 years. Only about half of respondents had submitted one in that time span, and only one-third had cited a preprint. Junior scientists were more likely to embrace these practices.

Among respondents not reading or citing preprints, the most common reason was concerns over quality. Among all survey takers, 42% predicted a strongly negative effect on science from preprints that spread misinformation. In comments submitted with their survey answers, some respondents voiced strong reservations about the growing use of artificial intelligence (AI). “Professors [could] mass-generate preprints with AI,” wrote an unnamed associate professor. These could “crowd out legitimate scholars who are publishing at a slower pace because they are actually doing real studies and going through peer review.”

Worries about quality may come disproportionately from clinical researchers concerned that the lack of independent vetting of preprints could jeopardize patient safety, says Richard Sever, chief science and strategy officer at openRxiv, a nonprofit that operates the widely used bioRxiv and medRxiv preprint servers devoted to biomedical science. (The new study does not report responses separately by subdiscipline.)

But concerns over quality may be based more on researchers’ impressions than evidence, Sever says, noting that bioRxiv and medRxiv reject submissions that don’t use the scientific method or that pose obvious risks to public health. Preprinting a fraudulent manuscript exposes it to more scrutiny than if it appeared only in a journal, he adds. “If you get a reputation for being the person who always puts up stuff [on preprint servers] which doesn’t have complete data and is shoddy, then you’re done in academia.” What’s more, some 80% of preprints eventually appear in peer-reviewed journals. And despite their quality checks, journals publish problematic papers, he says.

Respondents to Ni’s survey also saw upsides to preprinting, with about half agreeing it can accelerate the dissemination of scientific findings compared with journals, where peer review can take months and much of the content is paywalled. That finding echoes results of a survey of 7000 bioRxiv and medRxiv users, conducted by openRxiv in 2023 and posted on 26 February, in which respondents praised fast dissemination of findings as a top benefit.

Only about 16% of respondents agreed strongly that preprints reduce the importance that professional evaluators—those who review grant applications or make hiring and tenure decisions—place on articles in subscription-based, selective, peer-reviewed journals. Shifting away from traditional journals is a goal that advocates of open science have touted and some funders have embraced. For example, in 2025 the Gates Foundation began requiring grantees to post as preprints all manuscripts that result from research it funds, and it stopped paying for researchers to publish their papers in journals that charge a fee to make papers free.

Still, many universities’ professional review procedures explicitly prefer or require peer-reviewed publications, Ni notes. More than 60% of the survey respondents involved in funding, hiring, or tenure decisions said they give more credit to peer-reviewed papers than preprints; less than 12% said they credit both types equally. “Nobody has time to read preprints from 30 candidates for a position or award to determine their value,” an associate professor wrote in another survey comment. “Thus, we use journal [publications]. At least as a reviewer, we know there has been some bar surpassed.”

To help readers better judge the quality of preprints, Ni’s preprint suggests that preprint server managers find automated ways to summarize the rigor and transparency of each manuscript they post. Ng, who co-authored a 2024 survey of biomedical researchers’ views on preprinting, cautions that any such indicators “would need to strike a careful balance [to] avoid the oversimplification of research quality into a single score or checklist.” He argues professional evaluators need to judge the transparency and rigor of applicants’ research for themselves. “If institutions want to encourage open science practices, they need to ensure that researchers are not penalized, either explicitly or implicitly, for sharing their work early.”

Read More