Science is not perfect but it is important – that’s the point

In a recent Washington Post op-ed, Robert Gebelhoff suggests we recognize that science is far from unassailable and easily twisted to push various political agendas or to bolster our own particular world view. Gebelhoff then points to a variety of ways that science has been misused, and concludes by suggesting we refuse to have opinions where we don’t have a clear answer.

Whereas we should all be skeptical of proclamations from authority, Gebelhoff is far from alone in missing key aspects that give science its power. Let’s start by asking, what do you mean by “science”? Often nebulously imagined as a search for ultimate truth, science is no more nor less than the best tool we have for answering questions about the natural world. Ideally, those who do science abide by strict rules of conduct. Your hypotheses, laws, and theories must be testable, falsifiable, and predictive. Your claims are only valid if you have data to support them. You recognize your own inherent human bias and so have your work evaluated carefully by experts, a process known as peer review. If your peers and others cannot replicate your observations or experiments, the burden falls on you to either provide more evidence or reject or modify your conclusions. Ultimately, the data decide.

Science is a tool, but it is used by humans and that means that it will never be “anywhere near perfect” as Gebelhoff laments. This is certainly true, but ignores the fact that it is typically scientists themselves who catch errors and correct them. As is often said, science is a self-correcting discipline: if data keep piling up that contradict previous or current hypotheses, we reject the old and embrace the new because it is closer to what actually occurs in nature. Being truthful may not hold much currency in some human affairs, but if you lie and distort your data in science, sooner or later you will be found out and it will almost certainly mean the end of your career. Believe me, peer review is not for the faint-hearted.

Yes, science is not perfect. But this misses a crucial concept: there is no absolute certainty in science — there is simply probability.  As scientists we recognize that we are human and can only realistically deal in samples. For example, when it comes to climate change, we simply can’t have all the data on all the clouds, carbon dioxide, and local temperatures everywhere and always.  Therefore, we indicate that our data suggest certain scenarios are more probable than others. The higher the probability, the more confident one can be that the predictions will come to pass based on the data.

Here we get to the root of so many problems at the intersection of science and politics. Especially now, many of us have taken rigid political sides and we hide behind our identity bunkers, secure in the knowledge that our politics are the right ones. Recognizing the authority of science, but failing to understand where that authority comes from, we cherry pick and twist what we have read about science to aid our fight. Few people go beyond a news blurb, Facebook post, or tweet, and rarer still do we read what the scientists actually reported. In other areas of politics, we can easily deny the significance of scientific findings by turning the nature of science itself on its head: until we can be certain, we claim, there is no need for action. How often have you heard something like, “Let’s just wait until all the data are in, and then we’ll make an informed decision.” This is simply a way to deny significant findings and to wait forever, potentially putting us in harm’s way.

Perhaps the most frustrating assertion in Gebelhoff’s article is that since the U.S. government spends billions of dollars on non-defense research each year, “with so much money at stake, it’s simply unrealistic to expect all scientists to act purely for the advancement of knowledge.” These sorts of statements really hurt those of us for whom science is not just a profession but a vocation. First, Gebelhoff’s claim that “$70 billion” is spent on non-defense research is misleading. Yes, in 2016 approximately $70 billion was spent on non-defense research and development, but it is “folded into the budgets of more than two dozen federal departments and independent agencies, and there may be little or no distinction made between activities” (https://www.aaas.org/fy16budget/federal-rd-fy-2016-budget-overview). Furthermore, as an example, in 2015 the total U.S. government budget was approximately $1.1 trillion, and science funding accounted for only 3% or about $30 billion of that money. Second, the majority of the money is often spent on supporting and training students and postdocs, as well as (rightly so) public outreach and education. Most scientists, if they ever receive this highly competitive funding, are not rolling in cash. Only 20% of the grant requests the National Science Foundation receives are funded each year, and the average annual award is approximately $160,000. Think about that – the average annual grant award for supporting students, buying equipment, publishing the results for the public, and doing outreach is less than the average annual salary of a medical doctor in the U.S. The truth is, if you want to be a rich scientist you pursue a career in the private sector.

Gebelhoff is correct – science is not perfect and its conclusions can be twisted to justify political ends. But this is no reason to lose hope in the power and benefit of science as a tool. Recognizing that our politics often distort the nature of science, we must stop expecting science to give us absolute certainty to justify our preconceived notions. Instead, we must struggle to remember that science is an apolitical tool for understanding nature. It is powerful in that it helps us predict, often very accurately, the likely outcomes of our actions here on earth. Science doesn’t care about your politics, but like all tools, it can be bent toward noble or ignoble ends. Let’s choose wisely.

Yes, climate change is real. Yes, humans have played a role.

Today on the first full day of a new presidential administration that has already pivoted in digital media to downplay or possibly scrub climate change from the national conversation; and to join other scientists concerned with support of science including the science of climate change #USofScience, I am reposting my older post on science and climate change from a few years ago. The message is as relevant as ever. Climate change is real. The rapid climate change documented in recent history is tied to human activity. No amount of political argument or wishing it away can change that. I say that as a former climate change denier. I emphasize that climate data are what they are — carbon dioxide doesn’t care whether you are a Democrat, Republican, or Independent. Rapid climate change is occurring, and humans have played a role. What follows is my original and slightly tweaked post.


As the National Center For Science Education has been demonstrating for some time now, denying biological evolution and denying climate change are part of a larger phenomenon related to science illiteracy.   But I think we often tend to conflate the knowing of scientific data with knowing the process of science itself.  As a college professor, I can tell you that  smart students who know a lot about the natural world don’t always actually know the process of science.  In one of my first lectures to undergraduates in the introductory biology majors course, when I press them to define science, hypothesis, and so on, very few can.  And I have come to believe that our current societal issue with accepting science is a fundamental misunderstanding of the process, not simply a dearth of facts.

In my undergraduate days, I was a climate change denier.  That’s correct — I felt that the evidence was at best equivocal for global warming.  If you couldn’t prove it directly, how confident could we be?  In fact, I felt a good amount of the environmental “science” out there was nothing more than misplaced hysteria or political propaganda. For those who do know me and my political leanings, you are probably surprised.

So I speak from experience when I say that I understand the reservations among many people when it comes to climate change.  Ask any climate scientist, and they will never tell you with 100% certainty that their predictions will come to pass.  In fact, these scientists rely on models of climate, and those models are a hypothesis of reality, not reality itself.  Remember, I was a science major with aspirations of becoming a paleontologist, so my undergraduate self decided that if we couldn’t be certain, we shouldn’t go around broadcasting that it was the end of the world.  In my undergraduate head, the best science was certain, and that was why paleontology was so difficult — a lot of uncertainty.

So here’s the thing — a climate scientist can show you a lot of data (see below), and can tell you based on their expertise which are the most probable outcomes of current trends, but if you were my undergraduate self, you would not be convinced.

From Wikipedia Commons: “This image is a comparison of 10 different published reconstructions of mean temperature changes during the 2nd millennium.”

Whether or not my younger self (let alone my older self) was stubborn or simply a bit daft (probably both), I again point out a key feature in the thought process: if it isn’t certain, it’s not good science.

So, the assumption or implication that good science is certain is the first part of the puzzle.  The second part of the stubbornness by many of us to accept climate change or perhaps biological evolution is that we want evidence presented in a court room.  We want the TV show Law & Order, and we want the good lawyer to give us an iron-clad argument, or to show that our opponent is a lesser person, or to literally give us a smoking gun.  We are convinced that science works like this, and that the person with the best argument and evidence wins.  And most importantly, that the winner stays the winner.  Nothing can ever overturn the win.  Good science should be certain and win the day’s argument, for now and forever.

But of course, science has little or nothing to do with certainty and court room drama.  There is no certainty in science — there is simply probability.  Because a good scientist recognizes that we are only human, and we can only realistically deal in samples, we can’t measure every aspect of the known universe, and we certainly can’t have all the data on all the clouds, carbon dioxide, and local temperatures.  Therefore, a good scientist will never say they have “proved” something — rather, they will indicate that their data suggest certain scenarios are more probable than others.  The higher the probability, the more confident one can be that the predictions may come to pass.

It took a while for this concept to sink in with me.  It took graduate school and having to do science, and taking an excellent seminar from Professor Emeritus Ronald Toth at Northern Illinois University, that finally made science as a process click.

(As an important aside, much of my thinking as a scientist I owe to Ron — so the “smart” stuff I say about evolution and science are me emulating him.  My evolution podcasts and understanding evolution website are extensions [and I hope a sincere form of flattery] of Ron’s approach.  Thank you, Ron!)

That means, as someone who earned a B.S. in Geology with a Biology major, I had no real concrete idea about science as a process!  I am not surprised nor judgmental that many of our undergraduates, let alone the larger public, don’t understand this either — but this I believe is what needs to be most addressed.

Even if you do succeed in uncovering something new or accurately predicting a trend, there will always be new data. The complaint you often hear about science is how we keep changing our damn minds — we knew Pluto was a planet, or we knew that birds were not dinosaurs, or we knew that cholesterol was bad, and so on. But the process of science requires that one keep testing the hypothesis, and to incorporate new data as it comes in.  So we’re not changing our minds to tick you off — we adapting our models and our understanding of the natural world as more data come rolling in.

What I realized at long last in graduate school was that scientists speak in probabilities.  And when you think about it, we deal in probabilities all the time, and we make decisions based on those probabilities, and we are okay with that.  Every time you get in a car, there is a probability you will be in an accident … but you probably still get in that car.  Imagine if someone told you that unless you could 100% guarantee that no accidents would ever occur, it was pointless to drive.

Okay, but now for something more ominous: what about the probability that you will get sick if you ingest salmonella bacteria.  I have been sickened myself by this nasty “bug,” and many people have died from salmonella poisoning.  But there will always be cases where someone ingests salmonella or another pathogen and doesn’t become sick.  Now imagine a friend tells you that since every time a person has ingested salmonella they haven’t always become ill or died, we don’t have enough data to know whether or not it is truly deadly.  Therefore, wasting money and resources on preventing the spread of salmonella is not advisable because we can’t know with 100% certainty that everyone who ingests it will get sick or die.  This person would probably not remain your friend for long.

Probability in science works along this spectrum — from low to high odds.  Low odds: you will be hit in the head and killed by a rouge meteorite tomorrow.  High odds: the climate will continue to change, with an overall trend toward higher global temperatures.  Can we be certain climate will change in these ways?  Not 100%.  But the probabilities are high … and that’s why we should be concerned: the scientific predictions of increasing global temperatures suggest our world will change in ways that, if we are not prepared, will be devastating.  Of course, we could wait until we’re certain, and we could wait for the ultimate court room battle of the sciences … but if the probabilities are high, why wait?  What are waiting for?  Waiting for all the data to come in (which will never happen) and waiting for 100% certainty (which will never happen) is simply another way of doing nothing in the face of probable danger.

If you understand that the process of science is by its very nature is one based on probability, not certainty, I think we begin to get to the heart of the scientific illiteracy problem.  Giving people more and more data won’t help if they sincerely believe that uncertainty means no one knows anything.  This is, I believe, the core issue with science literacy — and why our politicians, our media, and our public are so often mislead to disregard good science and its important predictions that effect us all.

Evolution, climate change, and uncertainty: why understanding the process of science matters

As the National Center For Science Education has been demonstrating for some time now, denying biological evolution and denying climate change are part of a larger phenomenon related to science illiteracy.   But I think we often tend to conflate the knowing of scientific data with knowing the process of science itself.  As a college professor, I can tell you that  smart students who know a lot about the natural world don’t always actually know the process of science.  In one of my first lectures to undergraduates in the introductory biology majors course, when I press them to define science, hypothesis, and so on, very few can.  And I have come to believe that our current societal issue with accepting science is a fundamental misunderstanding of the process, not simply a dearth of facts.

In my undergraduate days, I was a climate change denier.  That’s correct — I felt that the evidence was at best equivocal for global warming.  If you couldn’t prove it directly, how confident could we be?  In fact, I felt a good amount of the environmental “science” out there was nothing more than misplaced hysteria or political propaganda. For those who do know me and my political leanings, you are probably surprised.

So I speak from experience when I say that I understand the reservations among many people when it comes to climate change.  Ask any climate scientist, and they will never tell you with 100% certainty that their predictions will come to pass.  In fact, these scientists rely on models of climate, and those models are a hypothesis of reality, not reality itself.  Remember, I was a science major with aspirations of becoming a paleontologist, so my undergraduate self decided that if we couldn’t be certain, we shouldn’t go around broadcasting that it was the end of the world.  In my undergraduate head, the best science was certain, and that was why paleontology was so difficult — a lot of uncertainty.

So here’s the thing — a climate scientist can show you a lot of data (see below), and can tell you based on their expertise which are the most probable outcomes of current trends, but if you were my undergraduate self, you would not be convinced.

From Wikipedia Commons: “This image is a comparison of 10 different published reconstructions of mean temperature changes during the 2nd millennium.”

Whether or not my younger self (let alone my older self) was stubborn or simply a bit daft (probably both), I again point out a key feature in the thought process: if it isn’t certain, it’s not good science.

So, the assumption or implication that good science is certain is the first part of the puzzle.  The second part of the stubbornness by many of us to accept climate change or perhaps biological evolution is that we want evidence presented in a court room.  We want the TV show Law & Order, and we want the good lawyer to give us an iron-clad argument, or to show that our opponent is a lesser person, or to literally give us a smoking gun.  We are convinced that science works like this, and that the person with the best argument and evidence wins.  And most importantly, that the winner stays the winner.  Nothing can ever overturn the win.  Good science should be certain and win the day’s argument, for now and forever.

But of course, science has little or nothing to do with certainty and court room drama.  There is no certainty in science — there is simply probability.  Because a good scientist recognizes that we are only human, and we can only realistically deal in samples, we can’t measure every aspect of the known universe, and we certainly can’t have all the data on all the clouds, carbon dioxide, and local temperatures.  Therefore, a good scientist will never say they have “proved” something — rather, they will indicate that their data suggest certain scenarios are more probable than others.  The higher the probability, the more confident one can be that the predictions may come to pass.

It took a while for this concept to sink in with me.  It took graduate school and having to do science, and taking an excellent seminar from Professor Emeritus Ronald Toth at Northern Illinois University, that finally made science as a process click.

(As an important aside, much of my thinking as a scientist I owe to Ron — so the “smart” stuff I say about evolution and science are me emulating him.  My evolution podcasts and understanding evolution website are extensions [and I hope a sincere form of flattery] of Ron’s approach.  Thank you, Ron!)

That means, as someone who earned a B.S. in Geology with a Biology major, I had no real concrete idea about science as a process!  I am not surprised nor judgmental that many of our undergraduates, let alone the larger public, don’t understand this either — but this I believe is what needs to be most addressed.

Even if you do succeed in uncovering something new or accurately predicting a trend, there will always be new data. The complaint you often hear about science is how we keep changing our damn minds — we knew Pluto was a planet, or we knew that birds were not dinosaurs, or we knew that cholesterol was bad, and so on. But the process of science requires that one keep testing the hypothesis, and to incorporate new data as it comes in.  So we’re not changing our minds to tick you off — we adapting our models and our understanding of the natural world as more data come rolling in.

What I realized at long last in graduate school was that scientists speak in probabilities.  And when you think about it, we deal in probabilities all the time, and we make decisions based on those probabilities, and we are okay with that.  Every time you get in a car, there is a probability you will be in an accident … but you probably still get in that car.  Imagine if someone told you that unless you could 100% guarantee that no accidents would ever occur, it was pointless to drive.

Okay, but now for something more ominous: what about the probability that you will get sick if you ingest salmonella bacteria.  I have been sickened myself by this nasty “bug,” and many people have died from salmonella poisoning.  But there will always be cases where someone ingests salmonella or another pathogen and doesn’t become sick.  Now imagine a friend tells you that since every time a person has ingested salmonella they haven’t always become ill or died, we don’t have enough data to know whether or not it is truly deadly.  Therefore, wasting money and resources on preventing the spread of salmonella is not advisable because we can’t know with 100% certainty that everyone who ingests it will get sick or die.  This person would probably not remain your friend for long.

Probability in science works along this spectrum — from low to high odds.  Low odds: you will be hit in the head and killed by a rouge meteorite tomorrow.  High odds: the climate will continue to change, with an overall trend toward higher global temperatures.  Can we be certain climate will change in these ways?  Not 100%.  But the probabilities are high … and that’s why we should be concerned: the scientific predictions of increasing global temperatures suggest our world will change in ways that, if we are not prepared, will be devastating.  Of course, we could wait until we’re certain, and we could wait for the ultimate court room battle of the sciences … but if the probabilities are high, why wait?  What are waiting for?  Waiting for all the data to come in (which will never happen) and waiting for 100% certainty (which will never happen) is simply another way of doing nothing in the face of probable danger.

If you understand that the process of science is by its very nature is one based on probability, not certainty, I think we begin to get to the heart of the scientific illiteracy problem.  Giving people more and more data won’t help if they sincerely believe that uncertainty means no one knows anything.  This is, I believe, the core issue with science literacy — and why our politicians, our media, and our public are so often mislead to disregard good science and its important predictions that effect us all.