from The Contemplative Mammoth, by Jacquelyn Gill ? ?
I notice it all the time– on Facebook, in the comments of a science blog, over family gatherings, or listening to a radio talk show. Someone, maybe you, is patiently trying to explain how vaccines cause autism, perhaps, or why so-called “anthropogenic” global warming is really just due to sunspots or some other natural cycle. Perhaps you are doing pretty well at first, making use of passionate, heart-felt rhetoric and well-timed anecdotes. People are nodding their heads in agreement, and perhaps you’re even changing someone’s mind.
And then a scientist joins the discussion.
The conversation tends to devolve from here, turning into a debate and (often) ultimately a debacle. Scientists are notoriously difficult to argue with– for one, they’re so sure they’re right! This is true of most people, though– and it’s probably true of you. What makes it especially frustrating to argue with a scientist is the jargon they use; if you don’t speak their language, you’re probably not going to change their mind.
I have created this handy guide to arguing with a scientist precisely for people like you! I’ve collected the most commonly used phrases and translated them into everyday English, so that the next time you argue with a scientist, you’ll not only better understand their arguments, but you might learn how to make yours better, too.*
Sample size: Sample size, often referred to by scientists as “n” (as in, “number”) is how many observations went into a particular study. In other words, if you are carrying out some research on whether donuts should be provided at meetings and you interviewed twenty of your co-workers, then your sample size would be 20 (or, “n = 20?). For scientists, sample size is huge (er, pun intended)– the bigger the n, the better. If you only asked two of your colleagues in your survey, you might randomly pick the two that hate donuts, and that would result in donut-less meetings! So getting the sample size right is important, which is why so many scientists and market researchers use a sample size calculator – to ensure their sample size is sufficient for their research. Scientists talk about sample size in arguments all the time, to convince you that they have more “data” (evidence) to support their claim than you do. For example, if you say that there is one study that proves that global warming is natural, but the scientist cites 10,038 studies, their n = 10,038 and yours = 1. You’re definitely going to need a bigger sample size to win this argument! Sample size could refer to the number of subjects in a test (like the number of interviewees in the donut study) or even the number of studies themselves (maybe you conducted two sets of interviews at two businesses). So why is more, well, more? Bigger sample sizes give you a better sense of the natural range your data might have (aka opinions on donuts), and you’re less likely to draw the false conclusion that people hate donuts because you didn’t ask enough people.
Made with the Einstein Chalkboard Generator (click for link).
Anecdotal evidence: This is related to sample size. An anecdote is a story– like that one about the time that Aunt Millie’s poodle ate four loaves of fruitcake and, well, you get the idea. Scientists are not fond of anecdotes, as a general rule. They’re not very, well, scientific (more on this later). Basically, your anecdote about your chain-smoking uncle who lived to be 98 is like having a sample size of 1, and it doesn’t hold up to the hundreds of studies that show a link between cigarettes and cancer. In fact, your anecdote isn’t even as good as a sample size of 1 in this case, because the anecdote is a story, and not a scientific study (remember, each study tends to have a lot of subjects, so really your sample size is closer to 0.001). Anecdotes are messy– they aren’t set up like proper experiments, with regulations and control groups. Your anecdote may be a really powerful story, and it may help to illustrate a point, but it won’t win an argument with a scientist. When a scientist says, “the plural of ‘anecdote’ isn’t ‘data'” , this is what they’re referring to: a story isn’t sufficient evidence to draw a strong conclusion. Scientific studies may have problems, but they’re still better evidence than a story (I may have just made up that story about Aunt Millie’s poodle, and you’d have no way of knowing!). If you want to convince a scientist, you have to show them real evidence.
“That’s not scientific!” A common critique from scientists is that something isn’t “scientific.” By this, they don’t mean that you didn’t come to your conclusion using test tubes and wearing a lab coat. Really, for something to be scientific, it needs to be done according to the Scientific Method– that is, you formulate a hypothesis (an idea about how the world works) and you test that hypothesis by experimenting or collecting data, and you repeat this process as much as possible until you better understand whatever it is you’re trying to learn more about. Obviously, this is tricky for a non-scientist to do, but you can limit yourself to arguing about things that are scientific! For something to be scientific, you have to be able to take a measurement of something– that is, it has to be “quantifiable” (which has the same root as the word “quantity”). Things like how much you love your cat or whether the Flying Spaghetti Monster made the universe are not quantifiable, because they can’t be measured. Therefore, they are not scientific. That doesn’t mean you can’t have real feelings or meaningful discussions about them, it just means that those discussions are not in the realm of Science. You can make an emotional, moral, or philosophical argument when debating with a scientist, but just be really clear that you’re using a different kind of criteria. For example, we can talk about whether it’s possible to clone Neanderthals to see if they had speech, and we can talk about whether cloning Neanderthals is morally and ethically right.
Peer review: Scientists just love to use peer review in arguments! You can show them all the well-written blog posts, internet surveys, magazine articles, interviews, and anecdotes you want, but scientists put much more stock in peer reviewed studies. By “peers,” they mean fellow scientists. This may at first sound like something of a Good Old Boys’ Club, where scientists sit around and pat one another on the back for yet another brilliant deduction. This is actually not the case! Scientists conduct experiments and collect data, write up that data into an article, and try to get that article published in a scientific journal (basically, a magazine for scientists that mostly only publishes the results of experiments). Many, many of these papers get rejected every day! Scientists can be very hard on one another, and often disagree on major ideas and finer details of methods, interpretation of data, or theories. Sometimes, experiments are just plain badly designed, or the conclusions are too strong and not supported by enough evidence. When a paper comes out in a journal after peer review, it means it’s been refereed by at least a few scientists in the field. Peer review can have its problems (you sometimes get an inappropriate editor, or perhaps easy or harsh reviewers due to luck of the draw), but that’s where sample size comes in. If you have a lot of peer-reviewed papers (a large sample size!), chances are that this effect evens out and you can come to some real conclusions.
To win an argument with a scientist, you need to cite peer-reviewed research. A website with a poll linking vaccines to autism is useless; anyone could have made up the data, for example, and your results are skewed to only those people visiting that website already (so, a specific subset of the population). With peer review, you at least know that the publication has undergone scrutiny by other experts. What’s more, once the paper comes out in a journal, scientists see it and continue the discussion by doing their own studies. If a scientist disagrees with a paper, they may even submit a letter to the journal explaining why. But with blog posts, or magazine articles– even well-written and well-researched ones– the rules are different. One problem is that non-scientists (maybe you!) often don’t have access to peer-reviewed articles (many are subscription-only), and so you can be at a disadvantage. But if you can cite peer-reviewed studies, or ask the scientist to, then your arguments will be much stronger. When you read a blog post or an article, check to see if they’re citing good sources, too.
Bias: When arguing with a scientist, you want to avoid using lines of evidence that come from what could be called “biased” sources–that is, sources that may have a vested interest in holding a different conclusion than that drawn from the evidence. For example, if you were trying to argue (still) that cigarettes don’t cause cancer, you wouldn’t want to back up your argument with a website run by a tobacco company. The tobacco company has an agenda (to sell cigarettes) and are therefore unlikely to share information that detracts from this agenda. “But what about scientists?” you might argue. “Isn’t the pursuit of science a sort of agenda itself?” Sure, and scientists do have their own biases (which could be the subject of an entire blog). However, at the end of the day, whether (most) scientists make money doesn’t depend on their research results. Publicly funded scientists (those who fund their work from research grants that come from taxpayer dollars) get paid to do research, not to come up with particular conclusions. Ultimately, peer-reviewed climate research is just more trustworthy than research paid for by an oil company, for example. If you want to win an argument with a scientist, be very careful about the kinds of evidence you’re citing, where it came from, and who paid for it.
Consensus: Scientists often use “consensus” as the ultimate argument-winner, and for good reason. Scientific consensus is the collected opinions of all scientists, and not just the one you’re arguing with. There can be one or two scientists who disagree (just like there are a handful of people who don’t believe the Holocaust happened), but if the vast majority of scientists have reached consensus, it means that there is so much evidence in support of an idea that it’s basically a guaranteed thing, based on state-of-the-art knowledge. Remember that scientists often disagree, have debates, argue, and may spend entire careers fighting one another. The reason scientists generally reach consensus with other scientists (and not, say, your UFO abudction group) is because other scientists also use the Scientific Method, publish in peer-reviewed journals, and are informed by some of the basic principles I’ve explained above. Consensus is a powerful thing, and in some ways it’s amazing that scientists ever reach it at all! In other words, if a scientist says that the consensus is that you are wrong, there is probably nothing you can do. Unfortunately, saying, for example, “well, the consensus among my abduction support group is that UFO’s did build the Pyramids!” isn’t going to sway them, because you probably a) don’t have peer-reviewed studies, and b) rely on anecdotes (see above).
If you’ve struggled with arguing with scientists, it’s understandable! After reading this guide, you may realize that scientists can have very different worldview than many people, due to their training and the tools they use– In fact, there are entire branches of study that are involved with understanding scientists’ outlook and how it’s changed through time. Having a different worldview can make scientists especially tricky to argue with, but it can be done (which is proven by they very fact that they argue among themselves all the time!). Scientists may seem like robots, but they in fact make mistakes, have emotions, and are biased, just like everyone else (including you!). Just remember that 1) behind the jargon, scientists are human beings, and most of them can really be very nice people, and 2) you don’t have to be a scientist to successfully argue with one!
*Insert many caveats about generalizations and oversimplifications here. This is meant to be an introductory guide, not a thorough treatise. Historians and philosophers of science will likely cringe, but the goal here is the edification of the layperson.