Can too much science be a bad thing? Growth in scientific publishing as a barrier to science communication.

There’s no doubt we live in an exciting time of innovation and discovery. With thousands of academic journals currently in print, more scientific research is published today than ever before. And the amount of scientific research being produced only continues to grow each year.

According to a search on PubMed – the free online database developed and maintained by the National Center for Biotechnology Information (NCBI) – scientific research has increased exponentially over the last century. The amount of research published in 2014 (514,395) was more than triple the amount published in 1990 (136,545), more than 100 times the amount published in 1950 (4,432), and more than 3,000 times the amount published in 1940 (153).

Figure 1: Number of scientific papers on PubMed (1900-2014).

growth in scientific publishing on PubMed
Note: PubMed search terms correspond to the fields of science included in the NSF Survey of Earned Doctorates and are as follows: biology, chemistry, astronomy, physics, mathematics, social science, psychology, agriculture, life science, and information science.

One can only hope this recent explosion in scientific research will bring us closer to unraveling some of the greatest mysteries in nature – mysteries in physics, medicine, and brain sciences, among others. And although we should certainly remain hopeful and optimistic about future medical and technological advances, we shouldn’t ignore what might at first seem an unlikely question – is this tremendous growth in scientific research an unqualified benefit to society? Is it possible to have too much science? And, if so, can too much science be a bad thing?

The Perils of Too Much Science

If you pay attention to the news, then you’ve probably heard a lot of discussion over the last few years about some of the problems plaguing modern science – problems such as widespread use of questionable research practices among scientists (see p-hacking), failed replications of findings that were once thought to be well established, and high profile cases of outright scientific misconduct and fraud. Furthermore, over the past decade, there’s been a substantial rise in the number of papers that have been retracted from scientific journals – more than can be accounted for by growth in scientific publishing alone.

These problems have prompted considerable debate in the scientific community, with some even going so far as to wonder whether science is broken. And although there’s no single cause to the current dilemma, pressure on individual scientists to achieve and maintain a high rate of scientific publishing is likely partly to blame. When scientists are evaluated based mainly on the number of scientific publications they produce, there’s ample incentive to cheat, because frequent publishing can garner rewards, such as tenure, promotion, or a job at a prestigious university.

And even though outright fraud is rare, pressure to maintain a high rate of scientific publishing can exert subtle influences that are no less damaging to the integrity of science. Being only human, after all, scientists are prone to the same unconscious biases present in all of us – biases that occasionally distort decision-making and motivate self-interest from time to time. And these are the sorts of insidious forces that are at work when a well-intentioned scientist, concerned about getting tenure or a promotion, runs multiple statistical analyses on her data and chooses to report only the ones that yielded anticipated results. Or when a genuinely curious researcher obtains unexpected results from his analysis and promptly engages in HARKING – hypothesizing after the results are known. Although less nefarious than intentional fraud or fabrication of data, each of these practices threatens the integrity of science by limiting transparency and introducing subjectivity and bias into the scientific process.

There’s no question these issues are important, and thankfully corrective actions are already underway to strengthen the scientific process. But the integrity of science is only one point of concern to be raised. A potentially thornier problem related to growth in scientific publishing is how to better manage, organize, and communicate all of the new scientific discoveries being published each year. And how to ensure society reaps the full benefit from these exciting advances in science.

Science Overload: A Problem for the Scientific Community

More scientific findings are published today than ever before, but is society reaping the full benefits of each new discovery? For that matter, is the scientific community?

These are difficult questions to answer for a couple of reasons. First, the value of a given piece of research is not always immediately apparent. This is especially true for most basic research, where immediate real-world application is usually of little concern. Nonetheless, basic science frequently paves the way for later research into more practical and applied issues.

Moreover, it’s not always clear how best to measure the potential value of research. The most obvious measure of value is the number of times a scientific paper is cited by other researchers. But consider that, in the field of psychology, the typical research paper goes mostly uncited over the course of the first 5-10 years following publication. But is most of this research uncited because it is unimportant? Or does this point to a breakdown in how new scientific findings are communicated among professional scientists? Could it be that a lot of very important research goes uncited because it merely goes unnoticed in the ever expanding pool of new science? The idea of so much research going unnoticed within the scientific community is a frightening possibility.

Yet certainly, some scientists make a concerted effort to stay on top of all the current trends and discoveries in their field, so as not to miss a single new paper. But with more and more research being published every year, this is an increasingly daunting task. For instance, it’s been estimated that in order to stay up-to-date on all the research related to just a single disease (e.g., breast cancer), a scientist would need to scan 130 different journals and read 27 papers each week, a task that could easily occupy 75% of each work day. Although research output varies between scientific disciplines, it seems hard to argue that, in general, it’s nearly impossible for a single individual to stay completely up-to-date on all scientific developments within a given field.

Science Overload: A Problem for the General Public

As challenging as it is for professional scientists, the task of staying up-to-date on new scientific research is infinitely more difficult for members of the general public – those who have little or no professional scientific training and no professional network of colleagues, graduate students, and post-doctoral researchers to help keep them in the scientific loop, so to speak.

What effect does the ever increasing growth in scientific publishing have on the average citizen’s understanding of science and the world? Does a greater level of scientific output contribute to greater scientific literacy, as one might hope? Or might it actually have the opposite effect? For instance, if members of the general public feel overwhelmed and inundated by too much science, might they simply stop paying attention? Could the recent explosion of scientific research ironically lead to a less scientifically literate and informed citizenry?

Before delving more deeply into this issue, let’s first consider just how much science the general public presently consumes – or, more specifically, how much science is presently communicated to the general public through popular media outlets.

What Percentage of New Scientific Discoveries Get Communicated to the General Public?

Arguably, the primary way average citizens learn about new scientific discoveries is through popular media – newspapers, magazines, websites, blogs, and podcasts, to name a few. But of all the scientific research that gets published in a given year, what percentage receives coverage on popular media outlets? And what percentage gets communicated to the general public?

Because I’m a psychological scientist, I’m going to focus exclusively on media coverage of psychological research. So keep in mind that although the findings reported below might serve as a useful benchmark, they by no means apply to all fields of science.

To try to estimate the amount of psychological research that makes it into the popular media, I conducted a search on Eurekalert.org for Social/Behavioral Science press releases. I restricted my search to press releases published within the past year under “Research News” and to those that contained the words, “psychology,” and “journal” (to try to weed out press releases for research from sources other than scientific journals, such as professional meetings). This returned 486 results for 217 journals, and for 158 of the journals I was able to estimate the total number of scientific papers that were published during the past year (22,482 papers and an average of 142.29 research papers per journal).

So how much psychological research was communicated to the general public in 2015? After some mathematical adjustments to the data (see footnote below), I estimate only a tiny percentage – about 4.43%.[1]

Is there a Connection between Scientific Literacy in the General Public and Rate of Scientific Publishing?

So only about 4.5% of all the scientific research published in 2015 made its way to the general public through press releases and popular media outlets, at least in the field of psychology. Moreover, we can assume only a fraction of this media coverage was actually viewed by readers. Again, we should be cautious about drawing overly broad conclusions based on an analysis of media coverage of psychological research alone. Indeed, coverage of other fields of science is probably generally superior to that of psychology. But unless media coverage of other fields of science is astronomically superior to that of psychology, it’s likely there is still plenty of room for improving how the latest scientific discoveries are shared with the rest of the general public.

On the one hand, perhaps this is encouraging because it suggests that the general public probably does not feel inundated by too much scientific news on a regular basis. On the other hand, such little exposure to science is worrying given the exponential rate at which new research is being produced. If roughly 95% of new scientific research goes unreported to the general public, then are average citizens getting left behind when it comes to the latest scientific discoveries and advancements?

Although recent surveys indicate that American Adults are actually more knowledgeable about science today than twenty years ago, there’s ample evidence that many Americans are out of step with the scientific consensus on a whole host of important issues. Issues that matter greatly today and that will only matter more so in the future.

For instance, according to a recent Pew Research Poll, the opinions of average Americans do not line up with the current scientific consensus when it comes to (1) safety of genetically modified organisms (GMO’s); (2) use of animals in research; (3) safety of foods grown with pesticides; (4) the role of human activity in climate change; (5) human evolution through natural selection; (6) dangers of overpopulation; (7) safety of nuclear power; (8) offshore drilling; (9) and safety of childhood vaccines.

According to the poll, the greatest level of disagreement between scientists and the American public is over the safety of GMO’s. Whereas a majority of scientists (88%) believe GMO’s are generally safe for humans to eat, most members of the American public (57%) believe otherwise. There is also a considerable gap between scientists and the American public when it comes to ideas about the causes of climate change. Whereas 87% of scientists believe the earth is getting warmer due mainly to human activity, only 50% of the American public believes this to be true.

Obviously, there are many reasons why people might believe something despite scientific evidence to the contrary – religious beliefs, political ideologies, and cognitive biases, to name just a few. But could another important factor be the rate at which new scientific discoveries are published each year? That is to say, is the general public especially in the dark on scientific issues that are currently experiencing rapid growth in scientific publishing?

To find out, I once again turned to PubMed. My goal was to survey the scientific literature related to each issue highlighted in the Pew Research Poll and then to estimate, for each issue, the rate of increase in scientific publishing over the past 20 years. Toward that end, I restricted my search to papers published after 1994 and to those with specific key terms or phrases (e.g., “genetically modified organisms” for the issue of GMO’s) in either the title or abstract. Then, for each issue, I calculated the standardized slope of the regression line that best described the rate of increase in scientific publishing between 1994 and 2014.

As shown below in Figure 2, there is indeed a moderate positive relationship between the rate of increase in scientific publishing on a particular issue and whether the American public disagrees with the current scientific consensus on that issue.

Figure 2: Relationship between rate of increase in scientific publishing and gap between views of scientists and the American public.

pew research gap and sci publishing_2
Note: Three issues from the Pew Research Poll were omitted because my search returned an insufficient number of papers for PubMed to generate a plot of publications per year. The three issues that were omitted were: offshore drilling, bioengineered fuel, and fracking.

In general, the higher the rate of increase in publishing on a particular issue, the larger the gap between scientists and the rest of the American public. This is important because if the rate of increase in scientific publishing on a particular issue reflects, in any way, the perceived importance of that issue within the scientific community then, in general, the American public is least informed about the most important issues being addressed by modern science.

Thoughts on Improving Science Communication

Although there’s variability between fields of science, in general we have reached a point where the amount of research being published each year is far too much for any one individual to keep up with and digest. Without improvements to the way we organize and synthesize new scientific findings, the increasing volume of new scientific research could eventually become a significant barrier to science communication, if it is not already at this point.

Scientific research has increased at an exponential rate over the past century. Yet despite the recent explosion in new research, only a tiny percentage of new discoveries are communicated to the general public through popular media outlets – about 4.5% in the case of new psychological research. And although some surveys suggest average Americans are more knowledgeable about science today than 20 years ago, there’s evidence that the American public is particularly uninformed about issues that scientists perceive to be among the most important areas of inquiry today – issues that have experienced a particularly rapid growth in scientific publishing over the past 20 years and that are likely only to increase in importance and urgency over the next several generations, such as climate change.

So what can be done to improve popular science communication?

Obviously, this is an extremely difficult question to answer, and it’s something scientists, educators, and science journalists have grappled with for years. So there’s likely no silver bullet solution to the problem.

Nonetheless, I suggest members of the scientific community must, at a minimum, work together to make it easier for average citizens to use new scientific discoveries to answer questions that are important to everyday life – questions about matters such as the safety of certain types of foods, the causes of climate change, and the effectiveness of interventions designed to treat mental illness, for example. Due to a variety of obstacles – such as poor public access to science (see paywalls), the growing volume of scientific research being published each year, and insufficient coverage of scientific research in the popular media – it’s presently too difficult for members of the general public to determine for themselves the current scientific consensus on a particular issue. As such, average citizens frequently have little choice but to defer to the assumed wisdom of experts. This is not to say that scientists and science journalists should not be trusted to honestly communicate science, only that members of the general public should be able to view and inspect the scientific data upon which a current scientific consensus is based. That important and potentially life-altering research should be communicated to the public any other way is ironic considering deference to authority is completely at odds with the core tenets of science.

All of this means scientists and science communicators must do a better job of communicating general scientific trends, rather than just the latest hodgepodge of individual scientific findings – an approach that, while fairly standard among popular science news sites, can create confusion among readers, particularly when the findings of individual studies conflict.

To better emphasize general scientific trends, science communicators will need to embrace new forms of media and explore new technologies that can enable efficient communication of an extremely large volume of scientific findings. Instead of focusing solely on the latest “groundbreaking” studies, we should present readers with a fuller depiction of all the relevant work that addresses a particular question, so as to place each new finding in its appropriate context. This may mean that we will eventually need to rely less heavily on traditional forms of communication, such as standard text-based narratives, essays, and literature reviews, in favor of approaches that are less story-like and more firmly grounded in actual data – approaches such as (1) text-mining and text analytics to look for big-picture trends in published research, (2) formal meta-analyses to organize, compare, and synthesize the findings of all relevant research on a particular subject, and (3) interactive data visualizations for presenting visually engaging depictions of the results from a large collection of comparable studies.

At a time when more research studies are being published than ever before and vast quantities of information are available on the internet to be mined, synthesized, and analyzed, we need to implement changes to popular science communication – changes that emphasize a data-driven approach to communicating research and that make it easier for members of the general public to synthesize a large volume of research findings. After all, most people understand that no scientific study is without limitations, and that each individual finding must be taken with a grain of salt. Therefore, focusing exclusively on new individual research findings likely does little to change minds, aid decision-making, and support scientific literacy in the general public. If we don’t do more to support a primarily data-driven approach that emphasizes communication of general scientific trends rather than merely the latest disparate collection of eye-catching findings, then arguably we are failing as science communicators and as advocates of an informed citizenry.

 

Footnotes:

[1] In actuality, my search yielded 1,185 results, but Eurekalert only returns the top 500 results in a given search. Therefore, because I was only able to access 41% of the total press releases from 2015, I adjusted my estimate of total publications from scientific journals by multiplying the total number of publications for each journal by 0.41. As such, my estimate of the percentage of psychological research that made its way into the popular media in 2015 is calculated as follows: {[Number of press releases from journals for which I was able to estimate total publications in 2015 (i.e., 408)] / [0.41*Total number of research papers from journals for which I was able to estimate total publications in 2015 (i.e., 9,220)]}*100.

 

Brian Kurilla is a psychological scientist with a Ph.D. in cognitive psychology. You can follow Brian on Twitter @briankurilla 

10 thoughts on “Can too much science be a bad thing? Growth in scientific publishing as a barrier to science communication.

  1. You failed to mention the increase in the number of journals more interested in publication fees than the quality of the science, the so-called “predatory journals.” There seems to be a rise in the number of these in offshore locations and serve the same function as diploma mills or term paper mills that allow questionable research to be published that would have a hard time being published in reputable journals, thereby muddying the water for those trying to find legitimate research results.

Leave a Reply

Your email address will not be published. Required fields are marked *