• NATURE PODCAST

When did people arrive in the Americas? New evidence stokes debate

Hear the latest science news, with Nick Howe and Shamini Bundell.

In this episode:

00:59 Ancient Americans

Two papers suggest that humans were present in the Americas thousands of years before many people have thought. We examine the evidence. Research Article: Ardelean et al.; Research Article: Becerra-Valdivia and Higham; News and Views: Evidence grows that peopling of the Americas began more than 20,000 years ago

10:44 Coronapod

We discuss the latest results from vaccine trials around the world, and controversy in the US as COVID-19 data collection moves out of the CDC. News: Coronavirus vaccines leap through safety trials — but which will work is anybody’s guess

24:38 Research Highlights

How being green makes things easy for some frogs, and how waves will be affected by climate change. Research Highlight: How frogs became green — again, and again, and again; Research Highlight: Extreme Arctic waves set to hit new heights

27:11 How can science improve?

A new book highlights some of the flaws of how science is done. We caught up with the author to find out his thoughts on how science can be cleaned up. Books and Arts: Fraud, bias, negligence and hype in the lab — a rogues’ gallery

35:54 Briefing Chat

We take a look at some highlights from the Nature Briefing. This time we discuss a puzzling new insight into the expansion of the Universe, and an update to Plan S that will allow open-access research to be published in any journal. Nature News: Mystery over Universe’s expansion deepens with fresh data; Nature News: Open-access Plan S to allow publishing in any journal

Subscribe to Nature Briefing, an unmissable daily round-up of science news, opinion and analysis free in your inbox every weekday.

Never miss an episode: Subscribe to the Nature Podcast on Apple Podcasts, Google Podcasts, Spotify or your favourite podcast app. Head here for the Nature Podcast RSS feed.

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.

Transcript

Hear the latest science news, with Nick Howe and Shamini Bundell.

Host: Nick Howe

Welcome back to the Nature Podcast. This week, evidence of humans in the Americas 30,000 years ago…

Host: Shamini Bundell

And exposing the flaws in the modern scientific method. I’m Shamini Bundell.

Host: Nick Howe

And I’m Nick Howe.

[Jingle]

Interviewer: Nick Howe

As we mentioned last week, our pandemic spin-off podcast Coronapod is now part of this show. So, if you just want to hear the latest coronavirus updates then you can skip ahead to 10:59. Coming up first, though, I’ve been looking into some stone tools and new modelling analyses that suggest humans may have lived in the Americas thousands of years earlier than many people thought. Now, the subject of when humans first arrived in the Americas is a hot debate for archaeologists.

Interviewee: David Meltzer

Well, it’s actually a long-debated topic as well. This is an issue that first arose in the Americas in the 1870s.

Interviewer: Nick Howe

This is David Meltzer, an archaeologist who investigates when humans arrived in the Americas. Over the past 100 years or so since, archaeologists have gone back and forth on when this might have happened. More recently, thanks to additional archaeological and genomic evidence, researchers are starting to converge on a window for when humans first stepped foot on the Americas.

Interviewee: David Meltzer

What we have coming out of archaeology is a number of sites that are sort of in the 15,000-years-ago range, and so that gives us a minimum age. We know that people were here by that date. The ancient genomics is pointing to a split between northeast Asian populations and those groups that would come into the Americas happening around 23,000-24,000 years ago, and that kinds of gives us a maximum age. So, somewhere between that 23,000-24,000 and 15,000 years ago is probably the window within which we’ve got people leaving northeast Asia, crossing a land bridge and coming into the Americas.

Interviewer: Nick Howe

Within this window, for many archaeologists, it’s most likely that people entered the Americas around 16,000 years ago. This is due to the prevailing well-established archaeological evidence, which is also backed up by climatic conditions. Before this time – between 16,000 and 24,000 years ago – it would have been difficult for humans to migrate across a land bridge from Asia due to the presence of giant walls of ice in North America. This was a time period within the last Ice Age, known as the Last Glacial Maximum or LGM.

Interviewee: David Meltzer

If you get to Alaska during the Last Glacial Maximum, you’re in a cul-de-sac, and so you’ve got basically two massive ice sheets that are blocking your way south.

Interviewer: Nick Howe

But this week in Nature, there are two papers that push the date of human arrival back thousands of years before 16,000 years ago. One has examined a cave deep in the desert of Mexico, known as Chiquihuite.

Interviewee: Ciprian Ardelean

The oldest dates we have are somewhere around 30,000 years ago.

Interviewer: Nick Howe

This is Ciprian Ardelean, the lead archaeologist examining the cave. 30,000 years ago would push back human arrival to well before the Last Glacial Maximum when ice blocked entry into North America.

Interviewee: Ciprian Ardelean

There were some clues that that cave had something in those layers belonging to the LGM that seemed to suggest humans, and I was very sceptical, but Chiquihuite Cave showed me that if you go deeper and deeper in your excavations without stopping when you hit the marker of the 14,000 or the 18,000, you may get lucky. You may get into something that’s been invisible for quite a while.

Interviewer: Nick Howe

With evidence of humans during the Last Glacial Maximum, Ciprian was encouraged to look beyond this barrier and dig deeper to older sediments. And by digging deeper, Ciprian found an array of unusual stone tools that he thinks that were made by people that inhabited the cave.

Interviewee: Ciprian Ardelean

Flaked stone tools made of a very peculiar material, which is recrystallised green limestone. So, they searched for this particular kind of stone around the cave, and they consistently picked for the greenish variety and used it to make very weird-looking artefacts that are oddly shaped. So, it’s not typical flint or obsidian you would find in many places in the Paleolithic.

Interviewer: Nick Howe

Now, this isn’t the first evidence that has suggested that humans have been in the Americas this early. There are other sites where there have been claims of human occupation more than 20,000 years ago, but these have been disputed by some archaeologists. Also, there are comparatively few sites of this age in the US or Canada so far. If we assume that people initially crossed into Alaska across the land bridge known as Beringia and then migrated south, we’d expect to find more traces of them. Well, that’s where the second paper in Nature this week comes in. It’s authored by one of the researchers who’s been investigating Chiquihuite, Lorena Becerra-Valdivia.

Interviewee: Lorena Becerra-Valdivia

So, we were looking at archaeological and chronometric – so that’s essentially radiocarbon dates and luminescent dates – from 42 archaeological sites from North America and Beringia.

Interviewer: Nick Howe

By collecting and analysing archaeological evidence from the associated dates, Lorena and her co-author were able to build a model of human dispersal as populations fanned out from Alaska across North America. Whilst this model doesn’t find new dates for artefacts, it uses known dates to suggest when people would have first reached the different sites across North America.

Interviewee: Lorena Becerra-Valdivia

We were able to see that humans were present in North America before, during and after the Last Glacial Maximum, but that human expansion didn’t actually occur until a lot later during a period of abrupt sort of global warming.

Interviewer: Nick Howe

This idea that human populations were low during the Last Glacial Maximum and then rapidly expanded once things got warmer is also backed up by Ciprian’s archaeological evidence from the cave in Mexico.

Interviewee: Ciprian Ardelean

As soon as you reach the layers in the deposits of the cave that date to about 18,000-16,000 years ago, the number of artefacts just triplicates in number.

Interviewer: Nick Howe

Ciprian also thinks this could help explain why there is so little evidence of humans during or before the last glacial maximum.

Interviewee: Ciprian Ardelean

It’s almost impossible to find them. You have such a huge chunk of our planet with just a few footsteps on it. That’s how the LGM looked like. You can barely call it a populated continent.

Interviewer: Nick Howe

But do these papers roll back the dates on when humans were present in the Americas? Well, it may be too early to tell. For David Meltzer, who you heard from earlier and who wasn’t involved in this research, when it comes to the cave, he thinks there are still questions to be asked about the tools that were found.

Interviewee: David Meltzer

Based on the radiocarbon dates that they have, and I think the radiocarbon dates look awfully solid, it appears as though this technology lasted for minimally, according to the dates, around 16,000 years. With a stone tool tradition that long lasting, one would expect it to have been far more widespread in the region instead of being localised to this one cave.

Interviewer: Nick Howe

David acknowledges that perhaps archaeologists just haven’t found other stone tools like this, and that could account for why they don’t appear to be widespread. He did have questions as well regarding why the stone tools don’t appear to have changed over the 16,000-year period that they were found. For him, this is quite strange. Ciprian, however, argues that maybe that isn’t so unusual.

Interviewee: Ciprian Ardelean

If you look at other places in the world during the Ice Age, the stone technologies did remain the same for many thousands of years without significant changes, so that’s what happens at Chiquihuite. It behaves more like stone tool industries in the old world where they remain unchanged for thousands of years.

Interviewer: Nick Howe

For the second modelling paper, David thought that Lorena’s analysis was sound, but he did have some questions regarding some of the sites that were included in the analysis.

Interviewee: David Meltzer

In some cases, the sites themselves, the data is highly ambiguous. So, you’ve got a well-dated site but the artefacts or the indications of a human presence may just be shattered bone or cut marked bone, and those of us that do field work that deal with this kind of thing know that there are a lot of natural processes that can mimic human actions on bone or stone.

Interviewer: Nick Howe

The debate surrounding when humans first came to the Americas is far from solved. These new papers add evidence to it but, in many ways, they raise a lot of questions as well. Perhaps, though, they will inspire archaeologists to dig deeper to find the first Americans.

Interviewee: David Meltzer

At the moment, subject to change, if you look at the sort of converging genomic and archaeological evidence, it looks as though people are coming in soon after the Last Glacial Maximum, so 16,000-15,500 years ago. Could they have been there earlier? Absolutely. But if you’re going to make the argument, it’s going to require a well laid out case.

Host: Shamini Bundell

That was David Meltzer from Southern Methodist University in the US. You also heard from Ciprian Ardelean from the University of Zacatecas in Mexico and Lorena Becerra-Valdivia from the University of Oxford here in the UK and the University of New South Wales in Australia. We’ll put a link to the papers discussed in the show notes.

Host: Nick Howe

Next up, Benjamin Thompson, Noah Baker and Amy Maxmen are here to give us the latest coronavirus updates in this week’s Coronapod. In the past, we’ve kept the podcast a coronavirus-free zone, so if you’d rather not hear this segment, you can skip ahead to 24:25.

Host: Benjamin Thompson

Yes, thanks, Nick. So, here we are again for Coronapod, and once again I’m joined on the line by Noah Baker and Amy Maxmen. Hi, both, how are you doing today?

Amy Maxmen

I’m good.

Noah Baker

Yeah, I’m not bad, thank you.

Host: Benjamin Thompson

We’ve got a couple of stories to cover today. The first one I thought we’d go for, well, there’s been a lot of interest this week in vaccines, and there’s been a mountain of data that’s been released in the past couple of days, really, about efficacy trials.

Noah Baker

Yeah, so this is something that we talked about on Coronapod back in May where we discussed some of the vaccine trials that are going on, and now there has been these big releases that have happened in the past sort of eight days. So, one is from Moderna, which is working with the National Institutes of Allergy and Infectious Disease – NIAID, I think, in the States is how you pronounce it. Another is from a group here in the UK which is from the University of Oxford. They’re working with AstraZeneca. And then there’s another being released from Pfizer, which is a US organisation that’s working with a German biotech company. And those three results have all come out and they’re all quite positive. They’re suggesting that, to different degrees and in different ways, there is an immune response to their vaccines in their stage II trials. Now, we kind of knew that this was sort of coming. From Moderna, for example, this was press released some time ago. But now, we’ve actually got data to look at, and that data always makes scientists happy, and it’s enough, certainly in two cases, to move on to phase III trials. The Oxford group have already started their phase III trial and Moderna are going to be starting their phase III trial very soon. So, all kind of exciting because it means things are moving and it means things are moving very fast, but what it definitely doesn’t mean is that we have an effective vaccine around the corner. This is just the next necessary step.

Host: Benjamin Thompson

Yeah, and there’s actually a fourth group too, the CanSino Biologics group in China, who have released some data, and they’re talking about doing a phase III trial as well. So, I think the prevailing thoughts here are that things are moving in the right direction, but clearly a long way to go just yet, and I guess it’s difficult to compare and contrast these vaccine trials and say which one is maybe the frontrunner because it’s a bit like apples and oranges in terms of their methods and their kind of analyses and what have you.

Amy Maxmen

For the phase III trial, that’s what we’re trying to see – efficacy. I think ideally there you have a place where there’s a lot of active transmission so that you can get a sense of if this is really protecting people. Is there any thoughts about the designs of those trials because the hotspots also move, so it’s super tricky? Do you know anything about that?

Noah Baker

Yeah, so that’s one thing that is definitely a problem that has been facing people doing these trials, is how do you try to do a phase III trial in a country that has all these sort of really, really extreme measures in place to prevent transmission because you sort of need there to be some natural transmission to be able to determine efficacy. And so, the Oxford group are running their phase III trial in South Africa and in Brazil, so they’ve deliberately chosen countries where there is a high transmission rate in order to run that trial. Whether or not that continues in the long term to be a place where there continues to be a high transmission, I don’t know the answer to that and I suspect they don’t either, but I suppose they have to choose somewhere and then run with it.

Amy Maxmen

It’s such a weird thing about vaccine trials for infectious diseases is that in some sort of sick way, you do better if the outbreak is worse.

Host: Benjamin Thompson

Yeah, here in the UK they have struggled to get people. I know they want to expand the range out to different ages but, you’re right, all these different lockdown measures that we have obviously encountered ourselves over the past few months, in a perverse way, are actually hindering potentially this particular vaccine being developed.

Noah Baker

Although we do have a sort of strong note of caution I think all of us are mentioning here because this isn’t a done deal yet. This is kind of incredibly fast. This is kind of unprecedented speed at which vaccines are being developed right now. I think the US group that’s looking into this is called Project Warp Speed and that’s exactly the reason is that this is very, very fast work that they’re doing. I think another thing that’s probably worth mentioning though is that when people hear ‘vaccine’, I think they immediately will jump to protection – ‘If I’ve got a vaccine, I’m immunised’ – but that’s not really how all vaccines work and, as of yet, we don’t know if that’s exactly how any of these will work. It might just mean that you have a temporary protection. It might just mean that you can still be infected but you just won’t get a severe disease, for example, so it will reduce the severity of that disease. The Oxford trial, early studies that they did in monkeys showed that their vaccine did reduce pneumonia, but it didn’t reduce the load of virus in the nose of those monkeys, so that kind of suggested that maybe they weren’t getting really ill but they were still able to pass that virus on. And so, there’s kind of still a lot of open questions about how any of these vaccines work, if indeed they do work.

Host: Benjamin Thompson

Yeah, I’m going to throw another spanner into the works as well, potentially. I mean just over the last couple of days, a preprint paper came out, so not peer reviewed, that suggested that levels of these neutralising antibodies drop off a cliff a few months after someone’s been infected with the virus. So, what does that means for the efficacy of a vaccine long term? Does it mean booster shots and what have you? It still remains to be see yet again.

Amy Maxmen

The FDA also announced, I think it was a few weeks ago, that they’ll accept rates as low as 50% efficacy, so they’re not saying it’s going to be great, necessarily.

Noah Baker

Yeah, I guess when you think about it, 50% efficacy doesn’t sound great, but if you put it in a context of versus nothing, that’s pretty darn good, isn’t it? It might be worth mentioning as well that just getting a vaccine that works, from sort of a biological and pharmacological perspective, is only part of the battle. The next part of the battle is also to produce it on a big enough scale and to get it to the right places at the right price. Now, those are big steps, and there are moves towards trying to do that. That’s why the Oxford group are partnering with AstraZeneca, and AstraZeneca have already said that they’re expecting to make 2 billion doses of this and do it at cost, so they’re doing it in a not-for-profit manner. And the same is true of an Indian pharma company which is intending to produce 1 billion doses of this vaccine but specifically for lower-middle income people, so priced at a lower level. But additionally, it’s not just a case of pharma companies committing to make loads of it. They also need kind of orders in place in order to be able to justify that cost, and so countries in the EU have already ordered something in the region of 400 million doses from AstraZeneca of the Oxford vaccine. But at the moment, it’s all kind of a bit of ‘we bet on the best data we have because we still don’t know if it’s going to work.’

Host: Benjamin Thompson

Yeah, the UK is backing quite a few horses, as I understand, and provisionally ordering doses from a bunch of different companies.

Noah Baker

And it’s also worth saying that there’s almost certainly not going to be one winner to this race. The likelihood is there will be a range of vaccines and a range of different approaches, and that’s not only going to be useful because they may all have different positives and negatives, but it’s also going to be potentially necessary because if you want to get enough vaccine to enough people, you need enough people working on it and that might mean more biotech companies or more groups that can work on it in order to be able to get it out and get it distributed at the right places.

Host: Benjamin Thompson

Well, let’s move on to our second story in this Coronapod then, both, and it’s about data collection. What’s the story here, Amy?

Amy Maxmen

So, the news that I think shook epidemiologists and public health researchers last week was an announcement that Health and Human Services – the US agency – has decided they’re going to make a new data management system for COVID where hospitals can report directly to them versus to the state health departments or the US Centers for Disease Control and Prevention. So, the CDC is a group that sits within HHS, but HHS is large and there’s like political appointees in charge of it. For example, the head of HHS, he’s not coming from a science background, and the CDC is an agency that was set up for public health. They’re duty is surveillance and collecting data from public health departments who they work really closely with. So anyway, this was really upsetting to a lot of people.

Host: Benjamin Thompson

Well that’s the set up then, Amy. What’s the issue with changing this to such a way?

Amy Maxmen

I don’t think there’s huge disagreement that we needed better data management because it was actually pretty disorganised and incomplete, but people were really upset for a few reasons. One is to switch to an entirely new system of reporting without any kind of leeway time for hospitals in the middle of a pandemic was really jarring for people. One other concern is so it’s one thing to say data management needs to be updated. We do need new systems. Hey, even if you’re going to bring in private companies, that’s fine. But why not do it in CDC? This is the agency that manages this sort of stuff, so why not modernise their system? So, nobody that I know has been able to give a good explanation for why HHS should be doing this and not the CDC.

Noah Baker

That is what I really want to ask about. This is the CDC’s job, right? If there’s a specialism that the CDC has, it’s how to do this. And there may be all kinds of problems with the systems they’ve got and they may not have been funded properly in the past or whatever it is, but it seems like a very strange idea to take this thing that the CDC are the specialist part of the government to do and give it to someone else and it’s not their specialism – why?

Amy Maxmen

And there’s not really a good answer for that, and that’s where people get really worried because of this history of what’s been going on with this pandemic and with this administration. There’s a lot of signs that the Trump administration has been really pushing its own narrative above scientific advice, so this is why people are really worried because if the people holding the data might be persuaded by politics while they’re analysing data, that could get really scary. And right now, we really don’t know how transparent this system is. Now, HHS has said they’re going to be very transparent. They have a new dashboard that’s up where you can see percent capacity full at hospitals, but that leaves out a lot of information and it’s kind of high-level data analysis. It’s not the granular level. So, they say they’re also going to be open with the CDC, so they have tried to reassure the public of that through press briefings but it really remains to be seen, and a lot of people are really sceptical at this point.

Noah Baker

What has the response from people in the CDC or epidemiologists in the CDC been to this? Are they outraged that they’re having this important sort of role taken away from them which has historically been their role? I can imagine I’d be a bit peeved if I was world-renowned for doing this and then someone said that someone else can do it.

Amy Maxmen

This is the state of affairs in America that I can’t tell you because try as I might to get CDC people and HHS people on the phone with me, those requests don’t go through and it’s been like that. So, they can’t tell me on the record how they feel about this, which is also, I find, very upsetting. But for a piece I’m working on I spoke with a lot of people who used to be at the CDC or who are former directors of the CDC. I talked with one of them and he told me this was completely unprecedented in his 30 years of public health. So, I think people who talk with people on the inside tell me that they’re surprised and upset about it, but I haven’t heard that on the record from anybody within the agency.

Host: Benjamin Thompson

When might any issues with this transfer then become apparent, do you think, Amy?

Amy Maxmen

I suppose we’ll know in the next like couple of weeks. To be fair, I want to give this the benefit of the doubt a little bit. Maybe I’ll be wrong and it will actually be a really great system. So, that’s not impossible. I haven’t seen that it’s manipulated right now. The day after this happened, there was a big newsflash when the CDC took down their data dashboard page, but from what I can tell that seems to be an error on the part of someone and what it does reflect is the confusion right now which is big. Nobody I know really can figure out exactly what’s going on so maybe it’s natural that there’s confusion right now, but I suppose you would expect in a couple of weeks we’d be getting good data out of this and if not then we’ll start to know.

Noah Baker

I think as well it’s probably worth mentioning that we have, on Coronapod and it has been talked about in Nature’s pages as well, the CDC hasn’t been playing a great game in this pandemic so far, and we do regularly criticise that we don’t have good enough data, and so in many ways, going the CDC haven’t been doing a great job, let’s do something radical to get better data, even if it’s in the midst of a pandemic and that isn’t necessarily the best way to do it, if it means we get better data and there’s political will behind it, in many ways, we might go, ‘Tick, they’ve done things we would approve of as an organisation.’ The question is whether or not that is the reason that this has actually been done and whether or not this change is actually going to work, basically.

Amy Maxmen

Yeah, like I said, it goes back to, ‘What’s the long game here?’ If there was a big overhaul of the data management system within CDC, that’s great for a lot of reasons because this is going to be the agency that’s going to continue to monitor this outbreak. This is going to go on for years, perhaps, so it’s going to be the one that’s going to continue to monitor this along with all sorts of other things that they collect data for already like hospital-acquired infections and antibiotic-resistant infections. Maybe that all plays into the things we monitor, so why put just this one box of data with different oversight? So, I’m all for data management but just a little bit worried about this switch.

Noah Baker

Yeah, it’s not like the CDC is losing all of its role, right? The CDC does still have a function. They’re not having the entire thing taken away. They’re not just being dissolved. It’s just this particular bit of data management is moving to somewhere else.

Amy Maxmen

Yeah, and it’s a big bit, and I should say, on the press briefing, Robert Redfield, the director of the CDC, said that he was informed of and approves of this change, so he didn’t protest it in public whatsoever.

Host: Benjamin Thompson

Well, let’s leave it there this week. I hope you’ll both join me next week for more Coronapod. So, for the time being, Amy and Noah, thank you so much.

Amy Maxmen

Thank you.

Noah Baker

Thanks, Ben.

Host: Nick Howe

And we’ll hear more from the Coronapod team next week. Coming up in a minute, we’ll be putting science’s flaws under the microscope. Before that, it’s time for the Research Highlights with Dan Fox.

[Jingle]

Dan Fox

Many frogs are green. Maybe not the cutting-edge research you were expecting, but unlike many animals, they don’t all have green skin. Instead, they’re green inside their bodies, and now, a team of researchers have found out why and how this helps them camouflage themselves so well. Tree frog skin is translucent, showing off their blood, bones and internal tissues that are all rich in the green pigment biliverdin. To understand why, researchers extracted lymph and other fluids from the polka-dot tree frog and traced the intense colouration to a previously unknown protein that binds to and transports the green pigment. Studying plants where the tree frogs live, they realised that the colour and brightness of the frog closely matched the vegetation. The biliverdin binding protein that allows for fine tuning of the amphibians’ colouration, perfecting their camouflage and allowing the creatures to vanish in the forest. Hop on over to the Proceedings of the National Academy of Sciences at the Unites States of America to read that research in full.

[Jingle]

Dan Fox

Climate change will swell the highest waves in the Arctic Ocean, endangering coastal areas and infrastructure. Researchers use five climate models to simulate how sea ice melts and wind pattern shifts might affect Arctic waves in the last two decades of the century. They found that in an extreme future scenario in which greenhouse gas emissions continue to soar, the maximum height of waves will increase by up to 6 metres off shore and by up to 3 metres along coastlines, with the biggest changes in the Arctic Ocean and Greenland Sea, as the loss of sea ice allows winds to build waves higher than ever before. Along the Beaufort Sea’s coastline in Alaska and Canada, the annual chance of damaging extreme waves would become four of five times more likely by the year 2100, and these extreme waves aren’t just a threat to coastal towns. Infrastructure such as offshore oil platforms are also at risk. Find the rest of that research at the Journal of Geophysical Research: Oceans.

[Jingle]

Interviewer: Shamini Bundell

Now, we’re all here because we are big fans of science – either that or maybe you didn’t check what the podcast was about before you pressed play. But being fans of science, we should be open to looking at its flaws as well its value, and many of those flaws are highlighted in a new book called Science Fictions: Exposing Fraud, Bias, Negligence and Hype in Science. I called up the author, Stuart Ritchie, and asked him why he became interested in this topic.

Interviewee: Stuart Ritchie

Well, actually, when I was doing my PhD, I ran a replication study of a famous controversial paper that had just appeared in a prestigious psychology journal, a paper that apparently showed that undergraduate students at Cornell University have psychic powers, that they could sense the future using extrasensory perception. We ran the same exact experiment again with a larger sample size and we found that we couldn’t replicate the effect. So, what we thought we would do is we would send the replication study to the same journal, but what we were told was that the journal does not accept, under any circumstances, replication studies. And it turns out that this story kind of stands in for a lot of problems in science where people are much more interested in new, exciting, flashy findings and less interested in the more boring kid of workaday findings which there’s not this sort of headline grabbing stuff but is much more likely.

Interviewer: Shamini Bundell

And the example that you’ve given of the study at Cornell, I mean you’ve made it sound sort of quite ridiculous, but these were esteemed psychologists. They clearly thought that they had found a real result in their data. How can you explain that?

Interviewee: Stuart Ritchie

So, there were actually nine experiments in the paper, and I’ll just describe one. He showed people two curtains on the computer screen, and they were told there was a picture behind one of them, and they were told to click the one that you think is the picture behind. And it turned out, if they picked a really boring picture behind then people get it 50/50, as you would expect. But if you put pornography behind one of the curtains, people are statistically, significantly more likely to choose that curtain, and there could be all sorts of reasons why they found that – a statistical fluke, there are various different sample sizes in the study which kind of might imply that what he was doing was collecting data and then if it wasn’t significant, adding more participants until it was a significant result. That’s possible. The statistical tests it uses are quite lenient, for instance, and various other problems. All of which together were not anywhere near anything we would describe as misconduct or fraud or that, nothing like that at all, but the standard practices that we use in psychology research and lots of other research too which make us more likely to get positive results but also make it less likely that the results we find are going to be replicated.

Interviewer: Shamini Bundell

And so, the kind of things that you’re talking about here are situations where, as you say, it’s not deliberate misconduct, it’s not someone trying to fake the results, it’s maybe a result of human psychology itself.

Interviewee: Stuart Ritchie

I mean first of all, I think we do have a big problem with deliberate outright fraud, but that’s a kind of separate thing from what happens much more commonly. I think there’s a much more common, and in some ways much more kind of insidious because it’s so widespread, problem of bias towards finding exciting, statistically significant results in the literature. So, if you look at the scientific literature, a huge proportion of the findings that are published there are positive results, right, way more than we would expect. In one study, it’s something like over 90% or maybe even more that. There’s also biases within the way we do the statistics, and this is what we’re talking about with the psychic powers paper, where there are ways of kind of leaning on putting your finger on the scale or whatever you want to call it, and the problem is that we can convince ourselves that these are absolutely fine, that it’s just you doing the usual kind of scientific exploration on your data. ‘Oh, I’m just checking it would work. I’m just trying it again. I’m just doing it.’ Rather than what’s actually happening which is that you’re massively increasing the chances of finding a false positive result, and the pressure that’s on us to find significant results and positive results is so much that we can really convince ourselves that any significant result is the big thing and it’s worth publishing.

Interviewer: Shamini Bundell

And the replication crisis in psychology has come particularly to the fore in the past few years, and I wonder if some scientists would have a tendency to think of this as maybe a social science problem and once you get into the ‘harder’ sciences, it’s much more factual and less open to personal bias.

Interviewee: Stuart Ritchie

Well, I think the first answer to that is that it’s probably better in ‘harder’ – if you want to call it that – sciences. However, there are lots of reasons to think that the hard sciences have serious problems with this as well. There are stories of replication failures. There was a survey done from Nature just a few years ago where a large proportion of chemists, for instance, said that they had real difficulty replicating results. In areas like machine learning and computer science you have real serious problems with people not publishing the code adequately enough for other people to come along and try and reproduce it. So, we have all these serious problems which, actually, psychology has done quite a lot of soul searching and kind of self-analysis itself, whereas in other subjects we don’t really know. We’ve just got these kind of glimmers, these kind of indications, but we don’t necessarily have the clear quantitative information, which I find actually, if anything, even scarier.

Interviewer: Shamini Bundell

So, in medicine, there have been some specific recent attempts to counter things like publication bias by, for example, pre-registering clinical trials, and in psychology, sort of attempts to go through and try and replicate things. What else has been going on in the last few years since more of this has come to light?

Interviewee: Stuart Ritchie

Yeah, you’re quite right. I should say that the journal that about ten years ago refused even to consider publishing our replication study now has on its website, ‘we consider replication studies,’ so something has changes in the past ten years and there are whole new ways to publish that involve you submitting your introduction and your method section to the journal and that gets reviewed before you collect any data, and the journal agrees to publish your paper whether or not the results are significant, so that’s a really nice new idea that’s come in. There’s also reforms happening at the level of funding, one of which is quite a radical idea which is that we should allocate funding for scientific research essentially via a lottery, which might stop scientists having to kind of over-egg the proposals.

Interviewer: Shamini Bundell

So, there’s lots of things that journals can do in these big institution changes. Do you have any advice for how an individual person can make sure they’re doing the best can?

Interviewee: Stuart Ritchie

What scientists can do is be as open and transparent as possible with their research. They can use a pre-registration protocol to write down their analyses before they do it, sharing your data with the world, putting it in an online repository, publishing in journals that encourage this kind of stuff, and generally not judging each other on how many publications we’ve got or in which journals they’ve published them, but judging each other on the quality of the research.

Interviewer: Shamini Bundell

And I suppose that has to be, to some extent, linked to a reform of how grants and funding is given in order for universities to be able to do that.

Interviewee: Stuart Ritchie

Absolutely. If funders just said, ‘Well, you have to put your data online or otherwise we won’t give you any money for it,’ then that would be a massive incentive for scientists to do the right thing. Maybe funding people to produce data in the first place, to produce resources that other scientists can use, is something that should be much more prominent in the way that funders are thinking about funding scientific research.

Interviewer: Shamini Bundell

Through the history of science, there’s also been the philosophy of science and people who have been very interesting in fine-tuning this scientific method over the centuries, and particularly recently people have become more aware of the kind of specific modern problems that we have. How much of a big problem for the world and for science do you think these things are today in 2020?

Interviewee: Stuart Ritchie

I think these kinds of issues of fraud and bias and negligence and hype and so on are probably worse now than they were a few decades ago because I think the clamour to publish or perish has got worse. I think there’s evidence of that. I don’t think these problems are completely new but I do think they’re a serious problem, particularly when you look at medical research which is really badly hampered by this. And then I hate to mention it because it’s all anyone can think about, but think about what’s happening in the COVID-19 pandemic. We just had two papers retracted, one from The Lancet and one from the New England Journal of Medicine, on the drug hydroxychloroquine, essentially because the scientists who were Harvard researchers had not looked at the data. They had relied on this data that had been given to them by this company, Surgisphere, and loads of questions have now been raised about that data and so the papers had to be retracted. So, I think a lot of the problems have just been exacerbated by this situation we’re in right now and I think it’s a real danger.

Host: Shamini Bundell

That was Stuart Ritchie from King’s College London in the UK, and you can find a review of his new book in this week’s Nature. We’ll pop a link to that in the show notes.

Host: Nick Howe

Finally on the show, it’s time for the weekly Briefing chat where we discuss a couple of articles that have been highlighted in the Nature Briefing – that’s Nature’s daily pick of science news and stories. Shamini, what’s hot in the world of science this week?

Host: Shamini Bundell

Well, I’m going to start off with a quick but very relevant shout out to my friend Amrit who’s Zoom background – you know you can do virtual backgrounds on Zoom – is the CMB, the cosmic microwave background radiation, which I thought was very clever as it is the ultimate background.

Host: Nick Howe

Laughs. Right, okay. I mean this is a charming story, Shamini. How does that relate to a story in the Briefing?

Host: Shamini Bundell

It’s about the CMB, obviously. I’m introducing the cosmic microwave background in a relatable fashion. So, do you know, Nick, what the CMB is?

Host: Nick Howe

It’s basically like the leftovers from the Big Bang, like some sort of radiation just permeates the Universe. It’s the static you seen in your TV, if that even happens anymore. I don’t know. How do TVs work these days?

Host: Shamini Bundell

Laughs. What’s a TV? Who still has a TV? I just watch everything on my laptop. Anyway, this is getting really off topic, sorry. So, yeah, close enough. The CMB was created back shortly after the Big Bang, and because of that, scientists use it to tell us about the origin of the Universe and how the Universe then is growing and shaping and evolving, and one of the things scientists are particularly interested in is how fast the Universe is expanding, and they put a figure on that which is called the Hubble constant. So, a few years ago, the Planck Telescope from the European Space Agency did some measurements on the CMB and came up with an estimated figure for the Hubble constant, which was all very well. But unfortunately, the figure that they came up with, based on the standard model of the Universe, didn’t match with direct observations of how fast galaxies seem to be moving away from each other in space.

Host: Nick Howe

So, I’m guessing then there’s some sort of new insight into the cosmic microwave background radiation and the expansion of the Universe?

Host: Shamini Bundell

Yeah, we’ve got new data. There’s been data published from the Atacama Cosmology Telescope, which sits in the Chilean desert and looks out into space. It’s also been mapping the CMB with amazing precision, and a lot of people sort of going, ‘Oh, yeah, this is great because then it will tell us whether maybe the Planck measurement was totally off and this will be much closer to what’s observed.’ But no, what the Atacama Telescope has shown is actually really close to what Planck said, showing that that initial value is probably pretty accurate yet doesn’t match other observations, so it’s sort of helpful and sort of just deepens this whole mystery of how fast the Universe is expanding.

Host: Nick Howe

Right, and are there any theories as to why these two numbers don’t match up?

Host: Shamini Bundell

Well, the article in Nature that I read didn’t put forward any specific theories, but one of the researchers was quoted as saying that they’re gut feeling is that there’s something interesting going on.

Host: Nick Howe

Sounds to me like maybe they don’t quite know yet, but hopefully we’ll find out soon. For my story this week, I’ve been looking into Plan S. This is a scheme that’s been put together by a lot of high-profile funders that has stipulated that any papers produced from the grants they give out should be published open access, and they’ve pushed different publishers to allow this. So, for example, Nature has said earlier this year that it will offer researchers a way to publish open access and be compliant with Plan S, but some other high-profile journals, such as Science, have yet to comply with the scheme. But in the latest in this story, the people behind Plan S have announced that it will allow researchers to publish in any journal, even if they haven’t complied with the scheme.

Host: Shamini Bundell

Well, that doesn’t make sense. If the journal is not an open-access journal then how can you publish in there and say it’s open access?

Host: Nick Howe

So, what they’re going to do is they’re going to allow researchers to post an accepted version of their article on an online repository as soon as it appears, so they’ll have the paper in the journal where they published it, and they’ll also have it on this repository too and then it will be open for people to see and read.

Host: Shamini Bundell

So, it’s a bit like a preprint server except you’re going to be hosting papers that have gone through the peer review and have been accepted by that journal. But surely the journal, if they’re not publishing it open access themselves, wouldn’t want anyone else to be publishing it open access either?

Host: Nick Howe

So, that’s sort of the problem. At the moment, only 20 publishers will actually allow authors to publish their paper on their site and also as an author accepted manuscript, at least in the way that is compliant with Plan S, which means that the paper can be adapted and shared widely. So, funders are going to basically override this prohibition so they can retain the right to share the manuscript regardless of what any publisher agreements say later on, and this will have legal precedence over any later agreement too.

Host: Shamini Bundell

So, the journal won’t be able to do anything legally about it being published elsewhere, but if the journal didn’t want their content being shared then they could also just not accept Plan S papers, right?

Host: Nick Howe

Yeah, exactly, so that’s sort of where the negotiations are at the moment, and for example, Science have now said it might permit people to publish in this way to allow some of the papers to be Plan S compliant in light of this new proviso that’s been added to the agreements.

Host: Shamini Bundell

Ah, so it’s a sort of get around that the journals don’t need to change their whole sort of system or setup but we can still have more open access work out there.

Host: Nick Howe

Thanks, Shamini, for chatting to me and listeners, we’ll put links to everything we discussed in the show notes, and if you’re interested in more but instead as an email delivered daily, then make sure you check out the Nature Briefing. We’ll put a link to that in the show notes as well.

Host: Shamini Bundell

Before we head off, it’s time for me to plug our YouTube channel again, as I always do. So, this week, we’ve got – warning, warning – coronavirus content, but it’s a video on basically six months into this sort of whole COVID-19 situation what do we still not know. What questions have scientists still got to answer. And we’ll put a link to the video in the show notes, so do check that out.

Host: Nick Howe

That’s all for this week. As always, if you want to get in touch with us then you can reach us on Twitter – we’re @NaturePodcast – or send us an email – we’re podcast@nature.com. I’m Nick Howe.

Host: Shamini Bundell

And I’m Shamini Bundell. Thanks for listening.

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing