Research integrity, AI, open access - how Elsevier is navigating the hot topics of scientific publishing

In this episode, YS Chi interviews Laura Hassink, Managing Director of Elsevier journals. 

Elsevier publishes 17 percent of the world's research output across its 3,000 journals. This gives it a unique vantage point on some of the biggest issues currently discussed in the research and scientific communities: research integrity, trust in science, AI, and the 'pay to read' and 'pay to publish' models. Laura discusses these issues head on. 

This episode is also available on YouTube at https://www.youtube.com/watch?v=YdfVzdzUKLQ

Transcript

Copy Transcript Copied

YS Chi: 0:00

The Unique Contributions podcast is brought to you by RELX. Find out more about us by visiting relx.com.

Laura Hassink: 0:12

But now, indeed, we have really accelerated and we're using a whole range of generative AI tools to check the manuscript upon submission on a whole range of what we call'hallmarks.'

YS Chi: 0:41

Hello and welcome to our podcast. I'm exploring today the topic of scientific publishing and research integrity. Elsevier publishes 17% of the world's research output, with its 3,000 journals has a unique vantage point on the debate around the'pay to read' and 'pay to publish' models, also more widely known as subscription or open access model. On the collaborative efforts to safeguard trust in science. To discuss these topics, my guest today is Laura Hassink, Managing Director of Elsevier journal. Laura, welcome.

Laura Hassink: 1:14

Thank you YS. Good to be here.

YS Chi: 1:16

Before we get going, how about you telling us a little bit more about yourself?

Laura Hassink: 1:22

Sure. I joined Elsevier 28 years ago as a graduate trainee. I wasn't in science, like many of my colleagues but I studied economics. Ended up at Elsevier. Thought I would stay three years, but here I am after 28 years and still enjoying it every day.

YS Chi: 1:42

I can't believe you've been with us 28 years. I remember when I first met you 20 years ago and you've had the entire career. There must have been something along the way, a moment, an incident or event that helped you really make your stay in single company through your career. Can you think of something?

Laura Hassink: 2:03

Oh yes, absolutely. Because, to be honest, I was really adamant that I would leave after a couple of years. But, yeah, it's really... a), it's the customers that we serve. I just enjoy that every day, but it's also the change that we're driving. In all those 28 years, there has been so much change in terms of customer needs, more industry wide developments, but also our ability to serve customers with the help of technology. Increasingly we have... every time we have also different solutions for some of the problems that our customers are facing. Never a dull moment. Every year there's change and new developments, and that's really what I enjoy.

YS Chi: 2:48

Well, you played lots of parts and we're going to get to that later on as we touch the subjects. How about we start with fundamentals? Since there are people who think they know how publishing works for us at Elsevier, but there's a lot that people do not know. How about to tell us, what is scientific publishing? What is the size of it to start with?

Laura Hassink: 3:10

If I focus on Elsevier. We receive more than three and a half million article submissions last year, and out of that, we published just under 25% so 720,000 articles. We do that in more than 3,000 journals, and we work with more than 36,000 editors and 1.7 million reviewers on an annual basis, to get all those articles peer reviewed and make sure that we give them that that quality stamp, and that they end up in our journals.

YS Chi: 3:47

And that represents what kind of market share?

Laura Hassink: 3:50

It's roughly a 17% market share in terms of research output, so article output. We also look at the percentage of citations, because that's an indication of quality. If we look at citations, then we are accountable for about 29% of the citations.

YS Chi: 4:09

29%. Almost double the number of articles, which means that your papers are significantly more relevant to other researchers.

Laura Hassink: 4:17

Absolutely. It really says something about our commitment to quality.

YS Chi: 4:22

And all this sits on something called ScienceDirect, which you at one time, actually managed.

Laura Hassink: 4:27

I did.

YS Chi: 4:28

Tell me about the extent of ScienceDirect. Why is ScienceDirect such a awesome platform?

Laura Hassink: 4:33

Yeah, it's really a top platform. As a researcher, you can't think of any other platform where your research is so visible. It's actually in the top 200 most visited websites globally. It has a very strong search engine optimisation position, so it's super visible. But also... there's a wealth of content, and what you see is that... it's all trusted content. It's all peer reviewed content. And just to give you a figure, last year, the content from ScienceDirect was downloaded 2.5 billion times.

YS Chi: 5:09

That's amazing. Given that there are, what, estimated about 30 million researchers around the world, including even junior ones.

Laura Hassink: 5:16

Yeah.

YS Chi: 5:17

That's a significant amount of usage. So, then tell me the business model behind it. We hear about 'pay to read','pay to publish', subscription, open access. Give us a one minute version of what that business model behind is.

Laura Hassink: 5:34

Yes. We have, indeed, two payment models, which we refer to as 'pay to read' or subscription. Or 'pay to publish', which is also referred to as open access. Pay to read is the more kind of traditional model that has been around longer, and that is actually where, as a customer, often an academic institution, you pay for a subscription to a particular journal or collection of journals. That's... to the whole of ScienceDirect, and then you have access to those articles. It also means that authors in those articles, they can publish for free. The costs are covered by the subscription payments. The other model that we have is 'paid to publish' or open access. There, the author pays an APC, an article processing fee, at the point of acceptance of the article, and then the article is available for free to everyone to read. We are actually agnostic. So there's different preferences amongst funders, institutions and researchers, and we serve both. customers. I always say, every author should be able to publish in our journals regardless of any right requirements they have from funders or institutions or preferences. That's why we are agnostic, and that's why we are offering choice.

YS Chi: 6:59

Let's talk about the second model that's more recent, you say. Pay to publish model.

Laura Hassink: 7:05

Yeah. Last year, we published over 250,000 open access articles. That's actually a 30% increase compared to the year before, that's a steep increase. We now have over 900 full open access journals. We have been launching quite a lot of full open access journals, and the reason for that is because we, where we saw that we were offering sufficient choice in our kind of our subscription titles, our hybrid subscription titles, we were not doing so on the open access side. We want to offer again... we want to offer that choice. That's why we have been launching quite a number of titles, to really provide that range in the different disciplines, quality tiers, etc.

YS Chi: 7:49

And so it's really demand driven.

Laura Hassink: 7:51

Yes, it is absolutely.

YS Chi: 7:53

Right, and where is that demand, particularly coming from, in particular field of science or particular geography or particular type of research?

Laura Hassink: 8:03

It's particular type of geography. What we see in Europe, there's a big appetite for open access. But also it's coming in parts of the United States, but actually we also see some pockets in Asia, although less overall. It also differs from discipline to discipline. There are, for example, disciplines where there's less funding in general, and there you see a lower appetite for open access. For example, around mathematics. It doesn't mean that there's no appetite for open access, but it differs from discipline to discipline.

YS Chi: 8:37

Right. Now as this volume grows, Laura, there's clearly concern that the quality control may be at risk, and the fancy word we use in the industry is research integrity. What can you tell us about your concern around research integrity? Then we'll deal with how we're tackling it next.

Laura Hassink: 9:02

We have very strict guidelines and policies around, the do's and don'ts around scientific publishing. We've always had instances where authors in particular were not kind of following those guidelines. That's indeed what we then refer to as a breach to our research integrity guidelines. We've always had that at a small scale. What we see now is that, what is happening in the publishing industry is similar to what is already happening in other industries for years. And that is that we see online fraud at a larger scale. I think that's inevitable, but it's something we need to deal with, because we have a major responsibility as a scientific publisher, to make sure that what we publish can be trusted, and that means that we have been, really been stepping up to indeed tackle that increasing issue.

YS Chi: 9:54

These issues may be easiest to understand for our audience, by giving one example of a fraudulent or dishonest publication.

Laura Hassink: 10:06

I'm happy to give an example, but I think I should also explain that there's a whole range of issue... research integrity issues. It's not just one particular kind of concern that we see. There's also different actors, and we identify them from ignorant authors who... often younger researchers who just don't know the rules of the road and make a mistake. Then we also see individuals. These can be authors, sometimes also reviewers or editors who are not ignorant and intentionally...

YS Chi: 10:44

Bend the rules.

Laura Hassink: 10:45

Bend the rules, sorry yes. And then all the way to... well, fraudulent companies who have a business model.

YS Chi: 10:54

So you're talking about organised crime, almost.

Laura Hassink: 10:57

Absolutely, yes.

YS Chi: 10:59

Okay.

Laura Hassink: 10:59

And sometimes these companies also referred to as paper mills. Clearly we treat these different, these three different actors in different ways. Then back to your question to give an example, one example, what we see happening. For example, where authorships are for sale.

YS Chi: 11:19

Yes, how about explaining that to people who don't know what it means to actually sell authorship?

Laura Hassink: 11:25

Yeah, so it's important as a researcher that you publish, because that advances your career. In some cases, there is unrealistic pressure on researchers to publish. Where they do not have either the funds or the time to actually do their own research and publish an article. Then there's now the option to actually buy an authorship, that means that these fraudulent companies, they sometimes, they produce an article which is fake, but sometimes it's a real article but then they add a fake author. Then actually, that authorship position is for sale, they try to sell it and that means what we see. That's difficult to detect. So that means, for example, that after submission of the article... of course, the article goes through peer review. It means different versions of the article. That can mean that in practice then, at some point in that peer review process, an author is added. As you can imagine, on some articles, we have sometimes 20 or 30 authors. To catch that with the human eye, that one author was added. It's not so easy.

YS Chi: 12:29

Then you use technology to catch those things.

Laura Hassink: 12:31

Absolutely.

YS Chi: 12:33

Let's talk about that technology. There are some probably basic stuff that we've been doing for a long time, but lately I hear that you're coming up with significantly more enhanced technology tools. Some may even call it AI tools.

Laura Hassink: 12:48

Yeah. Indeed, for a long time, we already had kind of the basic tools. For example, for years already ,we're checking all submissions on plagiarism, for example, and we have a tool for that. But now, indeed, we have really accelerated, and we're using a whole range of generative AI tools to check the manuscript upon submission, on a whole range of what we call hallmarks. Signals that are all an indication that there might be potentially something wrong with the manuscript. We're now rolling out these different hallmarks one by one, and make sure that all other article submissions are checked against those hallmarks, even before they are sent to our editors for peer review. To give you other examples, we're looking at, are the citations relevant? But we're also looking at authors, reviewers. Are they who they say they are?

YS Chi: 13:42

Right.

Laura Hassink: 13:43

There's a lot of different things that we check.

YS Chi: 13:46

Right. We've been doing data analytics for a very long time at Elsevier and these tools have been deployed. But I guess the availability of GenAI tools is just so easy to apply, isn't it, for mostly text based process that you have?

Laura Hassink: 14:02

Absolutely and I think that's also the benefit of being part of a company like Elsevier and RELX where, we have so much knowledge in house, which is really helpful now. Indeed, of course. I mean, our colleagues in Risk have been dealing with this for a long time. They're experts on this. So yes, indeed we have, we're in contact with them to learn from them.

YS Chi: 14:32

There's one other aspect of research integrity that I know you pay a lot of attention to, and that is the importance of editorial independence.

Laura Hassink: 14:42

Yes.

YS Chi: 14:43

And how that really is directly linked to the trust factor. Tell us a little bit about that.

Laura Hassink: 14:51

Yeah, editorial independence is extremely important. That means that we as a publisher don't take any decisions on which articles are accepted or not. That is really with our editors who are independent. That is for us, really important. That also means, for example, that where we apply this technology on research integrity, that we will make recommendations, but that the ultimate decisions are taken by our editors. We use, in general, we use our technology to make it easier for our editors and our reviewers to make decisions, but we will not make those decisions for them.

YS Chi: 15:29

Right. You give them the tools, but you don't make the decisions

Laura Hassink: 15:33

Exactly.

YS Chi: 15:34

Now, when things go wrong, what is the process that you and the industry uses to reverse what is being wrong?

Laura Hassink: 15:45

Yeah, that's a good question, because unfortunately, we're still not able to catch everything. We have been publishing articles in the past when we did not have the tools that we have now.

YS Chi: 15:57

Right. Retraction. Complete retraction from the records of science.

Laura Hassink: 15:58

Some of those articles have issues, where you have concerns. And only because we have better tools now, these Yes, and what you will see... and you also see that, not only come to light. That means that we take... every article that is brought to our attention, or that we kind of... where we with Elsevier actually, with all the publishers as well. You will ourselves, because we're doing a lot of investigations ourselves. Also on the on kind of now published content, because we have these better tools. Every article is that. Every article see more retractions. Already, you saw more rejections last where we have a concern, or where there is a concern, we will start an investigation. What is going on? What is happening? If we get complaints, are those legitimate? We will year and you will see more retractions this year. If you also kind of go back to the author to hear their point of view, and then again, based on that investigation. That's look into it, you'll see that most of those retractions are really thorough, and can take, in some cases, quite a long time. Then our editors will take a decision on what to do with that article, and it could be that it requires a correction really also from older content, and that's really because and in some cases, the decision is taken to retract that article. there's a bit of cleanup going on. In my mind, that is really important, because that means that, we are, were needed. We correct that scientific record, and that ensures that the readers can really trust what they find on our platforms.

YS Chi: 17:37

Right, right. One last question on this subject. Once an individual is determined by the editorial board that she or he has committed an error in the paper and the paper is retracted. Whose responsibility is it, Laura, that has consequence to the researcher? Because we don't have, from everything I know, a black list of authors that cannot publish. We don't do that. Do you think that the institution that houses them should discipline these people in their own institution?

Laura Hassink: 18:17

Well, depending on the on the mistake that was made. Again, there's no one size fits all. If it was unintentionally and it was, the author has apologised. It really requires a tailor made approach, to see how best to discipline. In some cases, it was an ignorant author who made a mistake. I don't think that it then justifies, putting that career at stake. There's definitely a responsibility for the universities, for the research institutions, also for the funders. I actually think it already starts earlier on in the process. I think there's a real responsibility, a), to educate people about do's and don'ts, but it's also about making sure that there's a healthy kind of research ecosystem. In some cases, I think the burden on researchers is huge. Their expectations are not realistic, and that is what results in this behaviour, in the increase in research integrity issues. I think we also need to tackle that, and that requires an action from universities and funders.

YS Chi: 19:24

Yeah, but I understand that it is not publishers duty to do so, nor the authority to do so.

Laura Hassink: 19:31

No, although, I do feel there's something that where we also have a collective responsibility. This is really an area where we work very closely with, indeed, research institutions and funders and policy makers and industry bodies to really work collectively on, educating, preventing, but also, detecting, when it went wrong.

YS Chi: 19:55

Right and I know how much energy you and many of your colleagues are involved in industry organisations to set standards, to create, a working guidelines that are consistent.

Laura Hassink: 20:09

Yeah, because that is really important.

YS Chi: 20:11

Yeah and you educate your editors, your researchers and our institutions about our guidelines clearly to them, so that they do agree with them and follow them. I know it's a herculean effort, but it's one that is definitely worthy of doing.

Laura Hassink: 20:28

Yeah, I couldn't agree more, yeah.

YS Chi: 20:31

We've been talking about the author side. Let's touch very briefly on the reader side. You've given lots of options, right? You said that you're agnostic to how they want to publish in which business model. What about then the reader side? In order to have true trust in science, we need to make sure that researchers can get access to read, prior research work on which they build further research. What do we do to try to do more access driven options for readers?

Laura Hassink: 21:02

Yeah, every payment model has kind of its limitations in terms of, for authors to publish or readers to read. We indeed have various initiatives to try to overcome those limitations. One on the reading side is, for example, we are a member of Research4Lif. We're actually a founding member. An initiative that provides developing countries with free or low cost access to our peer reviewed content. That is really important, and that is really an initiative that, for years already, really helps researchers in those countries to actually read our content.

YS Chi: 21:40

I mean, that has started back in 2002 with Elsevier Health, and I remember it starting only with health related stuff, and then agriculture related stuff, and then finally, environment related stuff. And then we made a decision to open all subject areas, and not only journals, but also books, and subsequently, all the tools. Which is amazing.

Laura Hassink: 22:07

Yes.

YS Chi: 22:07

Tell us a little bit about the usage of it by this community of country researchers that are in developing countries.

Laura Hassink: 22:15

Well, there is a lot of good research being done in these countries. We actually see that also coming back in kind of the articles they are submitting. It really enables them to actually do science which is really important. Often also, science which is important to that particular country.

YS Chi: 22:32

That's right. I mean, I travelled to Africa and see these institutions that would otherwise have zero access, and yet they get to read everything that the so called richer nations researchers can get access to, and at the same time, not delayed. I think it's terrific programme that Elsevier have started in the industry and I hope it goes on for forever. Well, Laura, you know that you and I can talk about publishing forever, having been in this thing for over 30 years. But I really appreciate the insights you have shared with us today. I do have one more question on a personal front, and then I'm going to try to wrap it up.

Laura Hassink: 23:11

Sure.

YS Chi: 23:12

Tell me, Laura, is there one moment of real happy surprise along your professional journey? Because not only have you been on the publishing side recently, but you were in publishing innovation side, transformation side, you were on the product side and many other areas across the company.

Laura Hassink: 23:32

What I continue to be amazed and really positively surprised is what the kind of collective thinking power in Elsevier. All the colleagues, the teams we have in Elsevier, what they can do in the short time frame. I have to say those, for me are really the kind of the happy moments just to see that. There's so much knowledge, so much commitment, and that is a real, genuine kind of commitment. But also that collective thinking power. That is something that I'm not only super proud of, but also which, continues to amaze me and makes me happy.

YS Chi: 24:04

Yeah, it's very well said. But just to remind you, Laura, you're part of that too. You're the part that brings in that teamwork, that innovation, the application of technology. I would say I walk away today with a couple of important points about the huge growth in publishing and how we're serving them with multiple models. The fact that it's really not a homogeneous group. There's incredible amount of variety that you have to address to differently, even within the same publishing process. Thank you so much, Laura, for being with us today, and I hope that you enjoyed this opportunity as much as I have.

Laura Hassink: 24:50

I certainly did. Thank you so much for the invitation, YS.

YS Chi: 24:54

Well, thank you for listening, and don't forget to tune in for our next episode.