2026年线下学术会议还有必要吗？Eiko Fried 谈 APS 大会的三个问题

In this piece, I will argue that conferences have become too costly to justify formats that mainly transmit information — talks and symposia. For these, asynchronous video recordings or papers are simply better. We should instead organize conference programs focused on formats for which physical attendance is uniquely suited for: hackathons, structured debates, and other co-production formats.

Why go to conferences? There are many reasons, and at least for me they have changed over the years. I would say there have been roughly three phases, each encompassing earlier themes, and adding on to them.

Phase 1, graduate school and early postdoc, were all about learning content. I went to as many sessions I could, took tons of notes, and occasionally asked a nervous question. Phase 2, postdoc and Assistant Professor period, was still focused on learning, but shifted more heavily towards presenting our work. Practising how to give decent talks, how to respond to feedback or criticism, how to engage with different (and also difficult) audiences to get a point across. Phase 3 now is the genuine pleasure of meeting close collaborators and friends, exchanging thoughts and ideas during lunch or dinner — or while seeing sights — and making meaningful experiences together that will be remembered fondly. Building something akin to a family in academia, people who can lift each other up and support each other, across the day-to-day boundaries (location, country, discipline) that constrain interactions to some degree. And joining some boring admin meetings but we won’t talk about that here. In summary, I used to go for content, then for presenting, now for people — and perhaps that evolution tells us something important about what conferences are actually good for.

The yearly APS conference (APS stands for Association for Psychological Science) has been the one I have attended most often in my career, nearly 10 times. The first time was in 2014; I still remember panicking at the airport because I had forgotten to get my ESTA but could do it at a machine close to the check-in desk at Brussels Airport. The photo above is from day 1 at APS in 2014—I remember the conference fondly¹.

I attended every year up to and including 2019, 2020 was skipped due to COVID, and then I went again until the second Trump term started. In May 2026, APS was held in Europe for the first time, in Barcelona, and it was so wonderful to see so many of my friends and colleagues again, including many US friends.

However, going back to APS after a few years of abstinence, and having attended many smaller conferences where many things go right — including SIPS, the conference for the Society for the Improvement of Psychological Science — it’s a good time to sketch what problems I see with APS conferences, and how we could tackle this.

This dovetails well with the call for feedback from the new APS president Dr. Davis-Kean in her recent piece, “Making APS Conventions Work for You.” In what follows, I’ll summarize three main issues I see with conferences, some solutions, and why they already align with Dr. Davis-Kean’s strategic vision.

Conference challenges

Access and costs

The costs of attending conferences are fairly large, and start with money. I’m a full professor now, with a budget of around 1250€ a year for attending international conferences given by Leiden University. This is not enough for APS, even if I sleep in a dorm with 12 beds and take the train. Booking a sensible hotel rather than the conference hotel, and paying yearly APS membership, the conference fee, and travel, APS will cost me at least twice my yearly stipend if I cannot cover it from grants. And if I’m struggling to pay for this with all the privilege I have as a professor in the Netherlands, obviously this will be much harder for many others. There are many other costs, too. Losing 4 days of work isn’t trivial for me these days, and there are severe environmental costs involved in international long distance travel. And of course larger conferences such as APS also compete with the preference of many researchers for smaller, more specific conferences, where there is a shared language and talks can be more detailed or technical and where you usually have closer ties to your colleagues.

Poor use of in-person time

Nearly all content of APS — or any scientific conference, really — is one person speaking to a room where people sit and listen². This does not work so well for me. In 12-15 minutes, the duration of a normal talk, I can get a ton more information out of skimming a preprint, focusing on the information I’m particularly interested in, with the possibility of digging into the methods section or getting an efficient answer to a question I may have that is likely described in a paper. If the format has to be a 12-15 minute talk, and the goal is information transfer, we have to admit that asynchronous video and papers are often better. Therefore, an in-person conference must focus on what cannot be replicated well online.

Weak intellectual interaction

Talks are usually followed by a few individual questions. I often find them interesting³, but there is usually not enough time for everybody to ask their questions. This is a bummer for the audience who spent time sitting through the talks, and also for the speakers who came into the room already knowing what they were going to present, so the real learning experience is really engaging with the audience.

Nearly all communication consists of dialogues between two people⁴. If person 3 has an idea regarding a comment someone made, or a thought to connect talks 1 and 2 with something the discussant said, this is usually very difficult to handle because there is no dedicated space for this. There are exceptions (panel discussions, addressing a group of researchers during a discussion), but this is the norm.

And in my 10 years of APS, I don’t recall many clear, well-grounded disagreements among scholars, which is one of the best ways to learn about a topic. And I don’t mean rudeness or being tough on a person rather than on an issue, but I mean an argument or debate that split the attendees into groups of roughly equal size. Don’t get me wrong: I have gotten into big trouble arguing for more civil debates, especially online …

… but I mean that we should invite good-faith arguments, productive and structured disagreements, and red teams. One of the best sessions I ever attended, at my very first conference ever in Berlin 2010, was a panel discussion of two primatologists — one very famous for his work in the lab, the other for his observational work in the wild. It was a pretty intense and fiery debate about the merits of controlled, experimental work contrasted with that of high ecological validity. And one of my favourite APS keynotes I attended was by Dr Loewenstein⁵ exactly because it was argumentative.

From presentations to co-production

APS, like most conferences, is very presentation-heavy right now. The most audacious fix would be to change many of the sessions we currently have — symposia — and replace them with different formats that make better use of in-person meeting time.

In my view, there are some scientific formats that merit getting a larger group of people traveling all around the globe to coalesce at a specific point and time to meet in person — formats that are much harder to do online. And these are co-production sessions in which researchers do something collaboratively, such as brainstorming sessions, hackathons, working sessions, structured debates with audience participation, and so on. I will focus on hackathons as an example here.

Hackathons, traditionally more common in areas where people build software or hardware, are sprint-like sessions or events in which interested researchers from all career stages collaborate intensively on a given project. And they’ve been a huge, hidden blessing for psychological science. Did you know that some of the most impactful projects and advances in psychological science in the last decade — including the Constraints on Generality statement, the PsyArxiv preprint server, the Framework for Open and Reproducible Research Training (FORRT), and the massive Psychological Science Accelerator project — all started as SIPS hackathons, small meetings of random peers who shared an interest to fix a problem?

In 2019, inspired by experiences many of us had made at SIPS conferences, where the motto has always been to try to stop complaining and start fixing stuff, we⁶ organized the first hackathon at APS, which was well attended. Such sessions are creative and collaborative, and even short sessions can lead to outcomes that can be followed up on later. At Leiden University, we are lucky to have the Lorentz Center which funds 1-week workshops that are essentially longer hackathons. You meet with a group of experts on a specific topic — we organized a session around measurement processes in intensive longitudinal data, with mostly early career researchers in 2024. There were 2 inspiring talks to get people to think about new topics, and a few workshops for learning hands-on skills, but we spent the vast majority of time in small subgroups. Those groups were created collaboratively on day 1, based on what people wanted to work on or learn about. And two years later, there are tons of outputs from these hackathons, including work I’m really proud of.

APS 2027

Hackathons, working sessions, and structured debates align very well with the vision sketched by Dr. Davis-Kean for APS 2027, in a number of ways related to the suggested Integrative Science Symposia (ISS) sessions, controversies and solutions, and scientific reports summarizing psychological issues.

ISS sessions — integrative, interdisciplinary symposia with a number of speakers — could line up specific questions or challenges that could be followed by a dedicated hackathon or working session to actually get work done improving the particular research area or topic, sketch out open questions, or develop a tool to create inroads for solutions. I would also love these sessions not only to remain highly interdisciplinary, but also to invite perspectives that are tougher to square up with each other, so real debates ensue.

Controversies and solutions — yes! A debate format where we have good-faith arguments, that sounds wonderful. It will also showcase early career scholars that criticism, or Organized Skepticism as Merton called it, is a crucial aspect of cumulative psychological science. We should celebrate fair criticism, but we need to practice handing it out and receiving it. No better place for this than a big conference. Bonus points if we get the audience involved!

Scientific reports summarizing psychological issues is a great initiative worth quoting in full here:

“Some conversations with the APS Board and editors have explored using the convention to focus on a single topic in psychological science, with symposium programming introducing it and offering different perspectives. We’d then set aside additional time for interested members to begin outlining a scientific report covering what we know, what we don’t know, what we need to know, and how we’ll get that information. The goal is a consensus document, shared with our scientific community as well as the broader public, that we would publish in one of our existing journals. What are your thoughts on a consensus report like this?”

I could not think of a more obvious topic for hackathons or similar formats. Overall, changing the question from “how do we get more people to attend in-person conferences” to “what program can we offer that requires in-person attendance” may be a good path forward.

Footnotes

1. Except for the sunburn. I cycled across the Golden Gate bridge in flip flops and had forgotten to put sun lotion on the top of my feet. You could see the white stripes on darker feet for 2 years after. ↩

2. I’m going to use APS as an example, also because it’s my conference and I am following Dr Davis-Kean’s call trying to make it work for me. ↩

3. Even some of those that start, as they often do, with “so this isn’t really a question but more of a comment”. ↩

4. “The mistaken belief that dialogue can mean only ‘conversation between two persons’ is from confusion of dia- (across, between) and di- (two); as early as 1532, trialogue appears needlessly for ‘a conversation between three persons’.” Ouch. ↩

5. I-frame vs S-frame paper. ↩

6. Kelci Harris, Lorne Campbell, Emorie Beck, Jessica Flake, Melissa Kline, and I. ↩