How does evolution explain complex mimicry?

Posted on December 20, 2025 by Fallacy Man

Evolution is, in my opinion, the most fascinating topic in all of science. It provides elegant, compelling, and enthralling answers for everything we observe in biology. It really is true that nothing in biology makes sense without evolution. Unfortunately, not everyone shares my passion for this topic, and evolution is often poorly taught and badly misunderstood.

Some examples of the comments people were making

As an illustration of this, I was recently wasting time on Twitter (I refuse to call it X) when I came across someone sharing an article from a few years ago about the discovery of a new species of rove beetle (Austrospirachtha carrijoi) that had evolved a specialized abdomen that made it look like the beetle was carrying a termite larvae “puppet” on its back. The puppet provided a disguise for the beetle and allowed the beetle to live inside termite tunnels and even be fed by the termites!

It’s a very cool discovery, and a fascinating example of the crazy stuff evolution can produce. However, the person sharing the article had a different take, commenting, “random mutation is a retarded explanation for this btw.” Many others chimed in agreeing with him and committing one straw man argument after another. As if that wasn’t bad enough, even many of the people who were commenting in defense of evolution were also badly misrepresented evolution. These comments predominantly followed two paths, either erroneously claiming that mutations are not random or doubling down and asserting that random mutations could create this given enough time.

The reality is more complex and fascinating, and both the original poster and many of the commenters* were missing three critical aspects of how evolution works:

  1. While mutations are random, natural selection is, by definition, non-random
  2. Evolution is “blind,” meaning that it has no direction or goal that it is trying to achieve
  3. Evolution doesn’t happen in isolation. Species co-evolve together.

In the post, I want to walk through each of these using the rove beetle as an example and try to give you a better understanding of how evolution by natural selection actually operates. To any young earth creationists who might be reading, as always, I ask simply that you hear me out and try to actually understand these evolutionary mechanisms. As I’ve written about before, I used to be a young earth creationist, and although I thought I understood evolution, essentially all of my objections to it were actually straw men. Put another way, in this post, I am not trying to convince you that evolution is the explanation for this rove beetle, rather I am simply trying to demonstrate that when properly understood, evolution does, in fact, offer a compelling and rational explanation (see my other posts for evidence that evolution is correct, e.g., here, here, and here).

*Note: some commenters did get it correct by pointing out that mutations alone are a bad explanation but mutations plus natural selection is a very good explanation.

 Mutations and selection

First, to be completely clear, genetic mutations are random. Drop any notions you have about mutations being guided, or consciousness affecting mutations, or mutations being biased towards beneficial ones. Mutation is a random process that happens during reproduction, and it (accompanied by genetic crossing over and the independent assortment) results in enormous genetic variation. Most mutations are neutral, a few are harmful, and a few are beneficial (see this post for more details on mutations).

Second, I completely agree that mutations alone would be a very poor explanation for the diversity of life that we see. Even with millions of years and millions of individuals mutating and reproducing, the odds of the exact right set of mutations occurring in the exact right order to create something like a rove beetle (all while avoiding harmful mutations) are astronomical, and when you apply that math to all of nature, it becomes wildly implausible. That’s why natural selection is so important.

This is one of the single biggest misunderstandings about evolution by natural selection: it is not random. Any time you hear a creationist scoff about something evolving “just by chance,” they are committing a straw man fallacy. Mutations are random, but that’s just step 1. The next absolutely critical step is natural selection, which is, by definition, not random.

In each generation, numerous mutations occur, most of which are neutral (those do evolve randomly via genetic drift), but every once in a while, one of them is beneficial. The individuals with that beneficial mutation survive/reproduce just slightly better than the ones who lack that mutation. As a result, they produce more offspring than the other individuals, and that inherently means that the beneficial mutation becomes more common in the next generation (more offspring = more copies of the mutation). Those offspring carry that mutation with them, and, just like for their parents, the mutation gives them an advantage allowing them to produce more offspring, which means even more copies of that mutation in the next generation. Each generation, that mutation becomes more and more common in the population all thanks to simple math. That’s it. That’s how natural selection works. There’s nothing mystical or atheistic about it: it’s just simple math.

Note that I am glossing over some complexities of inheritance that are irrelevant for the core argument, but in many cases it would take two copies of the mutation to have a benefit.

When a negative mutation arises, the process is the reverse. Individuals with that mutation produce fewer offspring, which means that the mutation becomes less common each generation (i.e., nature selects against it).

Note that this process is not random. Which mutations stay and which mutations go is determined by the effects they have, and they are “selected” for or against simply by causing the production of more or fewer offspring. This is critical, because it means species can accumulate beneficial mutations rather than accumulating mutations randomly.

Let’s use dice as an example. Let’s say you have 10 regular, 6-sided dice, and you want to get all 10 on the number 1. We’ll think of each throw as a generation reproducing, and each number as a mutation. The odds of tossing the 10 dice and getting all 10 to land on 1 (i.e., random mutations) are extremely low. In fact, they are 6^10 or 1 in 60,466,176. You could sit there throwing the dice for days and never get it. That’s the random mutation model, and I agree that it is absurd; so now let’s add selection into the mix.

Suppose instead, that every time you get a beneficial mutation (i.e., a 1) it is kept (in the same way that nature selects the beneficial mutations). So now, each time you throw the dice, you keep any 1s, then throw the remainder for the next generation. In this scenario, you’d actually get to a set of ten 1’s very quickly (go try it yourself if you don’t believe me).

Out of curious (and because I’m something of a nerd), I programmed a quick commuter simulation to try this and see how long it would actually take to get ten 1s, and on average, it only took 16.5 throws. We went from odds of 1 in over 60 million to averaging 16.5! That’s the incredible power of selection.

This is why the mathematical arguments against evolution fail: they are focused on mutations while ignoring the selection component. Once you add selection into the equation, it becomes entirely plausible to evolve something like a rove beetle with a termite puppet. The math works.

Evolution is blind

Now that we have cleared up the math, let’s look at the “blindness” of evolution. I really like the dice example I used above except for one important caveat: it gives the false impression that evolution is working towards some ultimate goal (like us trying to get all 1s). In reality, nature is not trying to accomplish anything. There’s no goal in mind. Each generation, the genes that result in the production of more offspring inherently get passed on to the next generation in higher numbers, while the genes that result in fewer offspring inherently are less common in the next generation. In other words, evolution is working one generation at a time.

Species evolve based on their current environment, which means that a trait that was beneficial in one generation can become detrimental in the next generation if the environment changes. Likewise, a trait that was being selected for one reason can get repurposed for something else if the environment changes or the right mutation comes along. Take the wings of a penguin, for example. For the penguins great, great, great, etc. ancestor, the wings were selected for flight, and evolution evolved them accordingly. Then, conditions arose that made being a strong swimmer more important than being able to fly, so evolution repurposed the wings into paddles for swimming. Again, there was no conscious process, it was simply that each generation, the individuals who had slightly better wings for swimming were able to do a better job escaping predators and/or catching food, which resulted in more offspring and more of the genes for improved swimming in the next generation.

Turning back to our rove beetle friend, there is a huge spectrum of insects (including many other types of rove beetle) that have evolved to raid termite and ant colonies, and they range widely from very ordinary beetles (and other insects) that get attacked constantly by the ants/termites to critters like the one in question that are so highly specialized that the termites accept them rather than attacking.

Many rove beetle species have an enlarged abdomen that sticks up over top them and looks something like a halfway point between a regular rove beetle and our termite puppeteer. What good is that blob? Glad you asked. Termites perceive the world largely through touch and chemical pheromones, and in some cases, that swollen abdomen seems to produce pheromones that help the rove beetle go unnoticed in the termite colony. So, selection has been acting on the abdomens to turn them into enlarged hormone factories.

Now we can easily imagine a scenario in which these pheromone-producing rove beetles do great when the termites sniff them, but once the termites start touching, they run into issues, because the beetles don’t feel like a termite (thus blowing their cover). Then along comes a mutation that makes their swollen abdomen slightly more termite-like. Perhaps it constricts at one point like a body segment or has a tiny protrusion like a leg. This makes it slightly less likely that the rove beetle will be detected, which lets it produce slightly more offspring, which results in that mutation becoming more common in the next generation. Then, just like with our dice example, beneficial mutations start to accumulate as each of them is selected and all negative ones are selected against, until eventually, we end up with this stunning example of mimicry. Keep in mind that the mutation doesn’t need to provide an enormous advantage. Any slight increase in survival/reproduction will be enough to shift the math in its favor**.

In that scenario, the swollen abdomen might not have been a very good mimic at first, but that was fine because it was being selected for pheromone production, not tactile mimicry. Then, later on, with the right mutation, evolution shifted course and started evolving a tactile mimic.

To be clear, I don’t know for sure that what I described is that path evolution took, and you can no doubt think of other equally plausible paths. My point is simply that these plausible paths exist, and it is entirely possible to get to this “final” stage of a highly complex mimic one simple step at a time.

For some more examples and discussions of this type of mimicry I recommend Parmentier 2000. Guests of social insects. In Encyclopedia of Social Insects. Springer.

 For more details on evolution being blind see this post.

**Note: there technically are cases where other factors do override mutations that are only minorly beneficial. These include things like large amounts of gene flow from other populations and small population sizes, which allow genetic drift to override selection. However, for most large populations, even a tiny benefit gets selected because that is how math works (more offspring = more copies of the mutation).

Species co-evolve

Finally, you may still be thinking that everything I have said is all well and good, except that there is no way for the process to get started because a half-formed mimic would surely be detected. Put another way, if evolution was the cause of this superb mimic, then it is inherently true that even a slightly less superb mimic would not do as well. So how could my hypothetical example of a mutation that causes a rove beetle with a swollen abdomen to get a slight constriction possibly be advantages? Wouldn’t the termites detect that right away?

The answer is simply that we are seeing the late stage of an evolutionary arms race that has been going on for a very long time. It’s like looking at a modern fighter jet and asking how a biplane could ever have been useful. The answer, of course, is that when biplanes were in use, they were going up against other biplanes, and we only got to modern fighter jets by different militaries constantly trying to outdo each other (here again, evolution is not consciously trying to do anything, but the same sort of arms race still occurs due to the math).

So, let’s back the clock up several million years and return to my example of a rove beetle who has a swollen abdomen for pheromone production, then gets a mutation for a constriction on the abdomen. At that stage, the termites would be naïve to that sort of trickery. They would not have the ability to distinguish even a poor mimic because they’ve never faced that problem. This would give our rove beetle a slight advantage, but it would also create a selection pressure on the termites to do a better job identifying mimics. So, termite colonies that have genes that make them slightly less trusting of the rove beetles with constrictions will do better, produce more offspring, which means more genes, etc. This puts the evolutionary pressure back on the beetles, resulting in a selection pressure for any further mutations to make them more termite-like, but every time the beetle adapts, the pressure flips back to the termites. Every time the beetles evolves better mimicry, the termites evolves better mimicry detection, and every time the termites evolves better mimicry detection, the beetles evolves better mimicry. Back and forth the two go for millions of years, each evolving in response to the other’s adaptations.

Isn’t that neat? I find it absolutely fascinating, and I have very distinct memories of learning about these arms races for the first time as an undergraduate (bats and moths were the key example there). Learning about this topic really opened my eyes and helped me to understand the true power and flexibility of evolution. I hope it has helped you as well.

Conclusion

I hope you can now see that evolution provides a rational, internally consistent, and compelling explanation for how something like a rove beetle that mimics termites could come into existence. Creationism, in contrast, lacks a compelling explanation. All creationists can do is shrug their shoulders and say, “God wanted it that way,” but that’s a cop-out, not an explanation. To those of us who want to understand the natural world, it is completely unsatisfactory (further it runs into all sorts of issues if you believe in things like Noah’s flood: how did that beetle and its termite host survive the flood, then find each other afterwards and make it all the way to Australia?). Evolution by natural selection is the only reasonable explanation, and you simply cannot understand biology without it.

The paper in question is: Zilberman and Pires-Silva. 2023. A new species and morphological notes on the remarkable termitophilous genus Austrospirachtha Watson from Australia (Coleoptera: Staphylinidae: Aleocharinae). Zootaxa 5336

Related posts

Posted in Science of Evolution | Tagged , , | 1 Comment

Vaccines are tested against placebos

Posted on December 2, 2025 by Fallacy Man

Lately, my Facebook page has been flooded with people insisting that  “no vaccine is tested against a placebo” (sometimes stated with additional qualifiers like “double-blind” or “saline placebo”). This claim, like so many anti-vaccer claims, is blatantly false. Nevertheless, I think it is worth looking at it more closely to better understand the tools available to scientists, the information they provide, and which tools to use for which applications. People often act as if randomized placebo-controlled trials (RCTs) are the one and only valid scientific approach and no other study design is even worth considering. In reality, placebo-controlled trials are great for certain applications, but they are far from the only useful approach, and in many cases, they are actually a highly inappropriate method.

saline placebo vaccine comment resposne

Vaccines are tested against placebos (including saline placebos)

It is simply not true that vaccines aren’t tested against placebos (including inert saline placebos). Anyone who says otherwise is either lying or willfully ignorant. Don’t take my word for it, go to Google Scholar or PubMed or any other scientific database with medical research and do searches like, “saline placebo double blind vaccine” and you will find literally thousands of papers. This sort of testing is standard as part of Phase II and Phase III clinical trials. Vaccines routinely go through RCTs before being released to the public (including COVID vaccines, btw, e.g., this RCT on Pfizer’s vaccine [BNT162b2; Thomas et al. 2020] or this RCT on Moderna’s vaccine [mRNA-1273; Baden et al. 2021], both of which used saline placebos).

It’s also worth briefly noting that among those studies you will find numerous trials where the vaccine either did not work or had serious side effects and, therefore, never went to market. Vaccines are stringently tested and ones that don’t pass those tests are either scrapped or modified then retested.

Other types of controls

Although clinical trials do generally use inert, saline placebos, there certainly are some studies that either use an older version of the vaccine or the vaccine’s adjuvants as the control (“placebo”). There are usually very good reasons for this, so let’s talk about them.

Let’s start with why scientists might use an older version of a vaccine as a control rather than an inert saline placebo. I’d like to begin with an analogy. Imagine you are a seat belt manufacturer, and you have developed a new seat belt design that you want to test. What would be the best control for those tests: no seat belt at all or the current standard seat belt design? Obviously comparing it against the current standard design makes more sense, right? You already know that the current design works. You already know its safety profile. You already know the level of protection it provides and how often it ends up injuring someone (seat belts aren’t 100% safe or effective after all). So, the relevant question is not, “how safe and effective is the new design compared to nothing?” The question is, “how safe and effective is the new design compared to the current standard design?”

The same situation is true for vaccines. If you develop a new version of a given vaccine, it may make more sense to simply compare it to the current standard of that vaccine which already has a known safety profile and level of effectiveness, rather than comparing it to a placebo (this is especially relevant given the ethical issues that we’ll get to in a minute).

Using adjuvants as the control group often occurs for similar reasons. Namely, it lets scientists isolate the effects of the active components of the vaccine compared to the adjuvants. This is useful both because the adjuvants have generally already been tested and have known safety profiles and because it means that if problems arise in the vaccine group, then scientists can narrow down which part of the vaccine is causing the problem and make appropriate changes.

I say this a lot on this blog, but scientists aren’t stupid. They spend a lot of time thinking about experimental designs and their designs have to pass the approval of ethics committees before proceeding. They try very hard to ensure they are using the best design possible to answer the question at hand. So, when you see something like a trial that used an adjuvant rather than an inert placebo, don’t throw up your hands and assert that the research is all nonsense. Rather, take time to carefully look at the study. Read the authors’ justification for the design, and look at how it fits into the broader literature. There’s likely a good reason for the study’s design.

Ethical issues with placebo-controlled trials

Despite everything that I have said so far, it is true that after the initial clinical trials, most follow-up studies on vaccines are not placebo-controlled, and there are several important reasons for this. One of the biggest is simply ethics.

Once a vaccine has been demonstrated to be safe and effective, it becomes highly unethical to do a double-blinded placebo-controlled trial where you are leaving half the participants (often children) vulnerable to diseases that you know you could prevent. It is totally unethical to play Russian Roulette with their lives. Despite what anti-vaccers like to claim, the benefits of vaccines are so extraordinary and so well documented that it is unethical to withhold them for an experiment.

Fortunately, scientists have other tools they can use, many of which are actually better than RCTs when it comes to detecting the conditions that people are often concerned about (e.g., autism).

Other useful study designs

Beyond the ethical issues, placebo-controlled trials are also problematic because they are expensive to run, difficult to maintain adequate controls over long periods, and are data hungry when it comes to rare events. That last point is the one that I really want to focus on here.

Sample size is obviously important in research, and the rarer the effect, the larger the sample size needs to be, but that quickly becomes a problem with RCTs. Imagine, for example, that an outcome in question has a background rate of 1 in 1000 and the drug being tested increases that rate to 2 in 1000. How big of a sample size do you need to detect that increase that with an RCT? Even without getting into the mathematical details, you should easily be able to convince yourself that it is going to require a rather massive data set to detect such a rare event. Even at 10,000 participants in each group (a truly enormous RCT) you’d only expect 5 positive cases in the control group and 10 in the drug group. That’s not a very big difference.

So, in those cases, a much better design is often what is called a case-controlled design. I explained this in more detail here, but essentially, what it does is flip the study design so that you start with a group of people with the outcome of interest (e.g., autism), match them with a group of people that is similar except that they lack that outcome of interest (e.g., don’t have autism), then you work backwards to look for the potential cause of interest (e.g., difference in vaccination rates between the two groups). This study design is great for rare conditions because you start with that rare outcome, rather than starting with a huge number of people and waiting to see if some eventually develop it. So even if the outcome only occurs in 1 in 100,000 people, you may be able to find enough medical records to let you do a meaningful study. Even a few hundred participants can be quite powerful for this design, whereas a few hundred people would be utterly meaningless in an RCT for rare conditions.

Although not the point of this post, I’ll note that this design has repeatedly been used to study vaccines and autism, often with large samples sizes, and the result has consistently been that vaccines do not cause autism (e.g., Destefano et al. 2004; Smeeth et al. 2004; DeStefano et al. 2013; Uno et al. 2015)

Now, you may be thinking, “but vaccines cause wide-spread harm, not rare occurrences!” In which case, you’re wrong, but, more importantly, we have a better design for more common conditions. Namely, a retrospective cohort study. Again, more details here, but in brief, this design generally follows more of the “standard” experimental approach, but it uses medical records, rather than actually giving patients anything (thus avoiding the ethical issues). So, you can look through medical records to, for example, categorize people into groups that did or did not receive a given vaccine (but are otherwise similar) then look at the proportions in each group that developed an outcome of interest (e.g., autism). The big advantages here are that you don’t have the ethical issues, the cost is lower than RCTs, and you don’t have to personally follow patients for years. As a result, you can achieve massive sample sizes of tens or even hundreds of thousands of participants, something that is almost never possible in RCTs. So, this design can be much more powerful than an RCT (for certain questions) simply because of the enormous sample sizes it allows.

Once again, several massive cohort studies have looked at vaccines and autism and consistently found that there is no association (e.g., Hviid et al. 2019; Madsen et al. 2002; Anders et al. 2004; Jain et al. 2015)

A lack of correlation is generally a lack of causation/you don’t always need placebos

There is one more critical topic that needs to be discussed here. Namely, study designs like cohort studies and case-controlled studies are much better at showing a lack of effect than they are at showing causation.

Causation is a hard thing to demonstrate (something anti-vaccers don’t seem to grasp). To confirm it, you need to not simply show that two things change together (i.e., are correlated), but rather that nothing other than causation can explain that correlation. This is the beauty of a randomized placebo-controlled design and why it is the “gold standard” for testing effectiveness. By randomizing the treatment across patients and administering a placebo (plus careful statistical analysis), you can control confounding factors so that you can be confident that the changes you see in the treatment group are being caused by the experimental factor rather than simply being associated with it. RCTs are, in most cases, the best way to establish causation (at least for topics like medicine).

Because they do not randomize and do not include placebo controls, case-controlled studies and cohort studies generally struggle to demonstrate causation. If they were able to very carefully match their participants and include appropriate covariates in their models, they may be able to strongly suggest a causal relationship, but there are serious limits to their interpretation. They are, however, great at establishing safety, because while correlation does not automatically indicate causation, a lack of correlation does generally indicate a lack of causation.

To understand what I mean, we need to talk about causation just a little bit further. Imagine that you show that X is positively correlated with Y. So, when X goes up, Y also goes up. Does X cause Y? Maybe, but it could also be that Y causes X or that some other variable is causing both X and Y. You’d need an RCT to tease that out.

However, suppose that X and Y are not correlated. So changes in X do not correspond to changes in Y. Does X cause Y? No. It’s an easy answer. At least within the statistical limits of the study in question, if changes in X don’t result in changes in Y, then X doesn’t cause Y.

What this means is that when large cohort and case-controlled studies find a total lack of association between vaccines and something like autism, that is actually really good evidence that vaccines don’t cause autism. Think of it this way, how could vaccines be causing autism if the group that received that vaccines doesn’t have higher autism rates than the group that did not receive the vaccine? (again assuming appropriate statistical design, case-matching, and within the confidence limits of the study)

To put that another way, if you are one of the people who likes to harp on the supposed lack of placebo-controlled vaccine trials (again, the do actually exist), then I want you to look at the large cohort studies and tell me exactly how you think a placebo would have made a difference. Look at studies like Madsen et al. (2002), which looked at records for 440,655 children who received an MMR vaccine and 96,648 children who did not receive an NMR vaccine, then compared their rates of autism (they were not statistically different). Exactly how do you think things would have been different if the people in the unvaccinated group had received a placebo instead of simply not being vaccinated. Do you think the placebo would magically have prevented them from developing autism? What would that have changed? A lack of placebo is simply not a valid criticism for this sort of study. If people who received the vaccine didn’t experience higher rates of autism, then that is good evidence that the vaccine doesn’t cause autism.

“But what about the entire vaccine schedule!?”

Finally, one last specific criticism I sometimes here is, “sure, individual vaccines were tested against a placebo, but no one has ever done a placebo-controlled trial on the entire vaccine schedule.” For once, that claim is at least true (to the best of my knowledge), but it is also a meaningless demand for an impossible test. That test would be completely unethical and also extremely difficult to pull off. It’s just not a plausible experiment.

There are, however, other approaches that scientists have used. For example, there is this massive cohort study of almost a million children that looked at numerous combinations of vaccines, comparing them to single vaccines (Bauwens et al. 2022). There are also studies that looked at the effects of antigen load (i.e., the number of antigens children are exposed to from vaccines; DeStefano et al. 2013; Iqbal et al. 2013; DeStefano et al. 2013). Other studies have looked at adding vaccines to the schedule or the effects of vaccines given together or spaced out (Olivier et al. 2008; Arguedas et al. 2010; Vesikari et al. 2010). All of these are different ways to examine combinations of vaccines.

Additionally, based on everything we know about vaccines (including studies like the ones cited above), there is simply no reason to expect serious harm from the routine schedule, and no matter how thoroughly something has been tested, there will always be things that haven’t been tested.

Imagine, for example, that someone actually did manage to do an RCT on the whole schedule and found a lack of significant side effects, anti-vaccers could then respond (and likely would respond) with things like, “well what about the whole schedule + GMOs, or the whole schedule while taking aspirin, or the whole schedule while watching 10 hours a week of TV” etc. There are endless possibilities, most of which are utter nonsense.

Don’t get me wrong, when actual serious, plausible concerns arise, scientists should (and do) take that seriously and test accordingly, but that’s simply not the case with these arguments about vaccines and placebo-controlled trials. Likewise, just to be 100% clear, there are side effects from vaccines (just like all real medicines). No one is saying that they are 100% safe, but they are very well-tested and serious side effects are extremely rare. Nevertheless, I (and all scientists) welcome improvements in vaccines to make them even safer. Unfortunately, instead of perusing those improvements, we are left doing things like looking at vaccines and autism for the 100th time as if one more study will somehow make a difference.

Statistical note: Throughout, you may have noticed that I use phrases like “within the statistical limits.” This is an important, albeit somewhat technical, caveat. In brief, it is never possible to prove a negative. There are several reasons for this, but the most important for the topic at hand is that it is always possible that there is an effect, but it occurs rarely enough that you were not able to detect it with the current sample size. So even if you have a sample size of several million participants, you aren’t going to detect something that occurs once every billion patients. So you can never, for any treatment, say with 100% certainty that there is no effect. However, if with large enough studies and proper statistics and designs (as has been done for vaccines and autism), you can confidently state that if there is an effect, it is so extraordinarily rare that for all intents and purposes, there is no effect. Stated another way, for something like vaccines and autism, we are as close to being able to state that “vaccines do not cause autism” as we are ever going to be. We have used so many giant studies, that we can state that if vaccines cause autism, it is such an extremely rare side effect that it is statistically undetectable.

Related posts

Literature cited

  • Anders et al. 2004. Thimerosal exposure in infants and developmental disorders: a retrospective cohort study in the United Kingdom does not support a causal association. Pediatrics 114:584–591
  • Arguedas et al. 2010. Safety and immunogenicity of one dose of MenACWY-CRM, an investigational quadrivalent meningococcal glycoconjugate vaccine, when administered to adolescents concomitantly or sequentially with Tdap and HPV vaccines. Vaccine 28:3171-3179
  • Baden et al. 2021. Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine. New England Journal of Medicine 384:403–416
  • Bauwens et al. 2022. Safety of routine childhood vaccine coadministration versus separate vaccination. BMJ 7
  • DeStefano et al. 2004. Age at first measles-mumps-rubella vaccination in children with autism and school-matched control subjects: a population-based study in metropolitan Atlanta. Pediatrics 113:259–266
  • DeStefano et al. 2013. Increasing exposure to antibody-stimulating proteins and polysaccharides in vaccines is not associated with risk of autism. J Ped 163:561–567
  • Dolhain et al. 2019. Infant vaccine co-administration: review of 18 years of experience with GSK’s hexavalent vaccine co-administered with routine childhood vaccines. Expert Review of Vaccines 19:419-443
  • Glanz et al. 2018. Association Between Estimated Cumulative Vaccine Antigen Exposure Through the First 23 Months of Life and Non–Vaccine-Targeted Infections From 24 Through 47 Months of Age. JAMA 319:906-613
  • Hviid et al. 2019. Measles, mumps, rubella vaccination and autism: A nationwide cohort study. Annals of Internal Medicine.
  • Iqbal et al. 2013. Number of antigens in early childhood vaccines and neuropsychological outcomes at age 7–10 years. Pharmacoepidemiology and Drug Safety 22:1263-1270
  • Jain et al. 2015. Autism occurrence by MMR vaccine status among US children with older siblings with and without autism. JAMA 313:1534–1540
  • Madsen et al. 2002. A population-based study of measles, mumps, and rubella vaccination and autism. New England Journal of Medicine 347:1477–1482
  • Olivier et al. 2008. Immunogenicity, reactogenicity, and safety of a seven-valent pneumococcal conjugate vaccine (PCV7) concurrently administered with a fully liquid DTPa—IPV—HBV—Hib combination vaccine in healthy infants Vaccine 26:3142-3152
  • Smeeth et al. 2004. MMR vaccination and pervasive developmental disorders: a case-control study. Lancet 364:963–969
  • Thomas et al. 2020. Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine through 6 Months. New England Journal of Medicine 85:1761–1773
  • Uno et al. 2015. Early exposure to the combined measles-mumps-rubella vaccine and thimerosal-containing vaccines and risk of autism spectrum disorder. Vaccine 33:2511–2516
  • Vesikari et al. 2010. Immunogenicity and safety of the human rotavirus vaccine RotarixTM co-administered with routine infant vaccines following the vaccination schedules in Europe. Vaccine 28:5272-5279
Posted in Nature of Science, Vaccines/Alternative Medicine | Tagged , , , , , , | Comments Off on Vaccines are tested against placebos

Don’t be intellectually lazy

Posted on March 25, 2025 by Fallacy Man

stick figure, read before posting, article, title, fact check blue and whiteIn recent conversations on this page, I have been struck by just how intellectually lazy science-deniers usually are. This is hardly a novel observation, but I think it bears discussion. I also want to note that this sort of lazy thinking is common in politics and countless other topics, and it is very easy to fall into these bad habits. Critical thinking is a skill, and like most skills, it requires practice. Being well-informed takes hard work. Blind adherence to biases and preconceptions is much easier than rigorous fact-checking and serious contemplation. We are all prone to cognitive biases, but if we want to have rational views based on evidence and logic, then we need to acknowledge those tendencies and fight against them. We need to be humble and acknowledge the limits of our personal knowledge and be intellectually diligent and honest. Blind denial of any information you don’t like is easy and seductive, but it is not rational or intellectually rigorous.

To illustrate what I mean by being intellectually lazy, I am going to use comments on a recent post I wrote about masks and COVID vaccines (as well as a few others from related posts), and I’m going to broadly categorize the lazy responses into three groups: blind denial, refusal (or inability) to cite evidence, assumptions and generalities.

Blind denial

The aforementioned article took me a long time to write. I wanted to be thorough and present a fair and honest representation of the scientific evidence. So, I spent many days reading the literature, fact-checking, and making sure that what I was saying was correct. I cited roughly 50 peer-reviewed articles (mostly meta-analyses and systematic reviews) and read numerous others while preparing that post.

Unsurprisingly, many of the comments were, shall we say, less rigorous. They can be summed up simply as, “I don’t believe it, and you’re a gullible idiot for believing it.” None of these people presented actual logical reasons why my arguments were wrong. None of them presented problems with the studies I was citing. Indeed, most of the comments were made by people who probably did not even bother to read my original post. They simply saw a title that disagreed with their views, so they automatically assumed I was wrong.

This is the extreme end of being intellectually lazy. Automatically dismissing any information that disagrees with your view is the very definition of being close-minded and it is the epitome of intellectual laziness.

One of the things that I say over and over again on this blog is that we must always be willing to question our views. We must always be willing to actually consider contrary evidence and carefully examine the possibility that we are wrong.

I want to be clear here, there would have been nothing wrong with someone saying, “I don’t agree with you because of the following specific problems with the studies you cited…” followed by actual issues with the studies and appropriate contrary evidence. I’m not saying that you have to blindly accept contrary information. Rather, I am simply saying that you have to give it a fair hearing. If you are being given legitimate evidence, then it behooves you to take it seriously and carefully examine it before deciding whether it is correct or incorrect.

So, to everyone who rejected my statements regarding COVID vaccines and masks, my question continues to be, why? What specifically do you disagree with? Why is the evidence that I presented unsatisfactory? Or, to put it another way, if all of those massive studies from around the world are not satisfactory, then what would be? What evidence would make you reconsider your position? Being intellectually rigorous means taking the time to ask yourself these sorts of questions. If you are rejecting something, ask yourself why, specifically do you reject it? Can you cite specific problems with it, or does your response consist of vague generalities (see section three).

Let’s take a step back from the topic of COVID and talk more generally for a second. Surely, we can all agree that being well-informed inherently involves a willingness to accept contrary evidence, right? How could you ever possibly know that you are wrong about something unless you are willing to look at opposing evidence when presented with it? Refusal to even consider contrary evidence creates a self-reinforcing view that is immune to logic. Blind denial is so much easier than a careful examination of the facts, but it is a trap that we must avoid at all costs. Be intellectually vigilant, not lazy.

Refusal (or inability) to cite sources

The next form of intellectual sloth that I want to discuss is a refusal to back up your claims. This is another topic on which I spend a lot of time on this blog. “That which can be stated without evidence can be dismissed without evidence.” Stated another way, the burden of proof is always on the person making a claim.

So, for example, when I claimed that COVID vaccines saved millions of lives, I backed up that claim with multiple large peer-reviewed studies. In contrast, the good people in the comments simply made their claims without providing supporting evidence, and when asked to provide that evidence, they refused, disappeared, or dodged. That is intellectually lazy.

At this point, many people respond with something like, “we aren’t all walking around with a stack of papers all the time.” My response to that is two-fold. First, if you are going to enter into a public debate, and especially if you are going to enter into a debate where the other side has already presented copious evidence, you do, in fact, have an obligation to have evidence for your claims. No one should take you seriously unless you can provide that evidence.

Second, (and more seriously) at least in my experience, that response is usually an excuse made by someone who doesn’t actually have real evidence. Multiple times now I have had a one-on-one conversation with someone where I kept pressing them for sources before they finally admitted that, “well, it’s just something I heard somewhere.” That’s a big problem, especially when the thing they “just heard” is something like, “all scientists are fundamentally wrong about basic facts in their field.” Likewise, people often try to dodge a request for evidence with something like, “just google it.” This is a cop-out response that, at least in my experience, almost always signifies a position built on sand.

So, if you find yourself unable to produce evidence for your claims, really ask yourself, “why do I believe this?” Have you actually seen reputable evidence to back up your claim, or is it just something you saw/heard on the news, facebook, youtube, etc. Have you verified that claim? Have you fact-checked it and traced it back to its original source, or is it just something that you believe because it sounded correct to you? If you can’t provide the actual evidence, then why do you believe it?

This is what I mean by being intellectually vigilant. You owe it to yourself to make sure there is actually a logical reason you hold the views that you hold, and if someone asks you for evidence, and you can’t produce it, take that seriously. Don’t be lazy and shrug that off. Be introspective about your views and get to the root of why you think a given thing is true or false. Does your view trace back to verifiable facts from legitimate sources? If not, why do you believe it? Why would you want to hold a view that isn’t based on actual facts and evidence?

Once again, this applies to far more than just science, and, in my experience, an inability to cite specifics is usually a sign of intellectual laziness where someone holds a view simply because it feels correct to them, rather than because they have carefully examined the evidence.

For example, a few months ago I had a discussion with a relative who insisted that a particular politician was “destroying the economy with their socialist policies.” I responded simply by asking them which policies specifically were destroying the economy. I asked them to name the pieces of legislation. If their view was actually based on evidence, that should have been a simple task. If they were actually basing that strongly held view on a careful examination of this politician’s policies and their economic impacts, it should have been quite easy to direct me to some specifics, but they could not give me any species. Instead, they hemmed and hawed and made excuses and cited their “personal experience.” The sad reality is that they were actually highly ignorant of this politician’s economic policies and were basing their views on misinformation and fearmongering. They had been fed misinformation by biased sources, and rather than fact-checking and testing the validity of those claims, they blindly believed them because they fit with their world view. That is what I mean by “intellectually lazy” and the dangers of that approach should be obvious.

Here again, it is worth being introspective. If, continuing the example above, you think that a politician has caused X, but can’t give any actually specifics of what they did to cause X and instead have to rely on vague generalities, then why do you have such a strong conviction on this issue? Where is your information coming from? Is it reliable?

Assumptions and generalities

This final category is quite broad but essentially consists of dismissing the evidence because of sweeping generalities that are based on assumptions.

By far the most common expression of this form of intellectual laziness is the “shill argument.” Many people dismissed the large body of evidence I presented because they assumed that all the studies were influenced by money from pharmaceutical companies. To be clear, conflicts of interest should be considered when evaluating studies, but that is not what these people were doing. Rather than actually checking the studies for conflicts of interest (studies always declare them) and carefully considering the evidence in light of those conflicts (when they existed), they were simply assuming that all of the studies were hopelessly compromised and all of the worlds hundreds of thousands of scientists had been bought off. In reality of course, many of the studies were not funded by pharmaceutical companies and had no conflicts of interest, and even when a conflict of interest is present, it does not automatically nullify the study. Rather it is simply another piece of information that has to be considered (see details here).

Conspiracy theories more generally also fall into this category of lazy thinking. They make sweeping generalizations based on assumptions that they cannot verify and often haven’t even tried to verify. They try to wave a magic wand and dismiss any and all evidence they don’t like by boldly proclaiming unverifiable statements rather than actually looking at the evidence. This is, of course, so much easier than actually seriously engaging in a topic, but only one of those paths will lead to an accurate, evidence-based view of the world.

This post has become something of a rant, so I will end it simply by reiterating the title: don’t be intellectually lazy. You owe it to yourself to examine your views and ensure that you are basing your positions on a careful consideration of the evidence. Being well-informed is hard work. It takes effort to fact check, verify, and examine contrary evidence, but it is vital if you want to have a realistic view of the world rather the going through life blindly.

Posted in Rules of Logic, Vaccines/Alternative Medicine | Tagged , | Comments Off on Don’t be intellectually lazy

Masks and COVID vaccines were huge successes; ivermectin and hydroxychloroquine were not

Posted on February 27, 2025 by Fallacy Man

COVID19 presented an unprecedented challenge for modern science/medicine. Faced with a novel infectious disease, doctors, scientists, and health officials rose to the challenge in remarkable speed. That rate at which we went from never having heard of COVID19 to having safe and effective vaccines continues to astound me and is a testament to the power of the scientific method. Dealing with this pandemic was made all the more challenging by a politically-motivated anti-science campaign, with high-ranking government officials spreading one conspiracy theory after another and dangerously misleading the public at every turn.

What really concerns me, however, is that now that the dust has settled and COVID has shifted from a pandemic stage to an endemic stage, many people (possibly even an increasing number of people) are still under the delusion that the scientists got it wrong and the conspiracy theorists got it right. I constantly hear people say things like, “well fact checkers said the COVID vaccines were safe and look how that turned out,” or “’scientists’ said we were crazy for taking a dewormer, but now studies show we were right” or “I’m not opposed to science, but I am opposed to the way it is politically weaponized like it was during COVID.” Indeed, one poll found that a third of Americans think that the COVID vaccines killed thousands of otherwise healthy people.

These sorts of beliefs and comments are completely out of touch with reality and contrary to the facts. They are also extremely dangerous, not only because recommendations like staying up to date with COVID boosters continue to be relevant, but also because this type of thinking affects how people respond to other scientific issues and future disease outbreaks. This erosion in the public trust in science is incredibly damaging and based entirely on lies and conspiracy theories.

Therefore, I want to set the record straight on four of the biggest COVID topics that I see coming up over and over again: vaccines, masks, hydroxychloroquine, and ivermectin. In all four cases, despite popular perception to the contrary, it was the scientists/health officials who got it right, and the conspiracy theorists who were dangerously wrong. That’s not to say, of course, that every recommendation was correct or that every study produced helpful results. There were obviously missteps as scientists and health officials did their best to update their knowledge and recommendations based on new findings and the progression of the pandemic. Nevertheless, the overarching effect was the scientist’s/doctor’s recommendations saved millions of lives, while the conspiracy theorist’s claims were entirely bogus and cost lives.

As we go, I will be relying on peer-reviewed studies (i.e., claims backed by data), but I wanted to ensure that I am citing studies that are truly representative of the literature. Thousands of studies have been published since the start of the outbreak, and there are a large number of small, low-quality studies (particularly from early in the pandemic). Therefore, I will be focusing largely on systematic reviews and meta-analyses, especially reviews and meta-analyses of randomized controlled trials. These studies are considered to be the highest form of scientific evidence, because, when done correctly, they systematically collect a large number of studies and pool their results to look for a consistent pattern. As I have explained before (here, here, and here), on any well-studied topic you will find lots of outliers, lots of low-quality studies, and lots of statistical noise. Systematic reviews and meta-analyses try to cut through that noise to look at the over-arching effect and avoid cherry-picking studies. If a result is real, you should see that consistently reflected in the highest quality studies.

Face Masks

Let’s start with the use of masks for helping to control the spread of COVID. I will acknowledge at the start that on this topic, more than any of the others, there was initially a lot of confusion and mixed messaging. It wasn’t initially clear if masks were effective, and shortages in supplies made many governments want to save masks for those who needed them most. However, after those initial growing pains, most government health agencies around the world endorsed the use of masks (especially N95s), with governments often instituting mask mandates. From a disease-prevention standpoint, that was the right call.

Multiple systematic reviews and meta-analyses have confirmed that wearing a mask helps to control the spread of COVID, ultimately reducing the number of cases, hospitalizations, and deaths (e.g., Ford et al. 2021; Baier et al. 2022; Hajmohammadi et al. 2023). Indeed, so many of these studies have been published that one research group put together a systematic review of 28 systematic reviews (SeyedAhmad et al. 2023), with the conclusion that masks are beneficial at controlling COVID. This is extremely compelling evidence. Nevertheless, simply listing a bunch of studies is clearly unsatisfactory, and each study had its own end points, populations, and caveats, so let’s look more closely.

First, it’s worth noting that not all masks are created equal. Several meta-analyses and systematic reviews have shown that N95 masks are best, followed by surgical masks, with more mixed results on the effectiveness of cloth masks (Lu et al. 2023; Wu et al. 2023; Zeilinger et al. 2023). This is an important factor to bear in mind both when evaluating studies and when making your own choices regarding masking.

Second, you may be wondering about the contexts of these studies. Some of the studies cited so far were, for example, on healthcare workers, and while they clearly show a benefit to wearing masks in that setting, it is fair to wonder if that also applies to other settings. While masking is more important for the protection of certain groups, there is ample evidence that is effective in many settings. Unsurprisingly, there is a strong protective effect among the elderly living in assisted care facilities (see this meta-analysis: Chen et al. 2024), but there also appear to be benefits among younger groups. For example, Viera 2024 found that 10 out of 14 studies on masks in school settings found a protective effect. It is also worth noting that 2 of the studies that did not find a benefit were rated as having a “serious risk of bias,” and in many cases it was not clear what sort of mask or face covering was being used.

Looking at the effects of masks for the public more generally becomes messier from an experimental design/control standpoint, but multiple studies have still found ways to tackle it. For example, Rader et al. 2021 used survey data from 378,207 people to estimate mask use across communities, which they then correlated with community transmission rates. From this, they concluded that a 10% increase in mask use was associated with increased transmission control (i.e., wearing a mask reduces the spread of COVID).

The most compelling evidence for a general community protection effect comes from studies of mask mandates. A systematic review of these studies found that 21 out of 21 studies documented “benefits in terms of reductions in either the incidence, hospitalization, or mortality, or a combination of these outcomes” (Ford et al. 2021). Let’s look closer at some of these studies.

In many cases, mask mandates were patchy, with some local areas enacting them, while others did not. This allows for comparisons of areas with and without those mandates (ideally while controlling for other differences between them). For example, Huang et al. (2022) took data from hundreds of counties across the USA that implemented mask mandates and paired them with nearby, similar counties that did not impose mandates. They then controlled for environmental and economic factors and compared the counties before and after the mandates. Prior to the mandates, counties that would and would not eventually implement mandates had similar case loads, that were increasing at similar rates, but after the mandates went into place, the groups diverged, with significantly lower case rates in the counties with masks. At the start of the mandates, both groups were experiencing ~14 new cases per day per 100,000 people, whereas 30 days after the mandate, counties without masks had climbed to ~20 new case/day/100,000 people, while counties with mask mandates were at ~14.5 cases/day/100,000 (i.e., with masks the rate remained largely stable, while without masks, it continued to climb). By 40 days, counties without mandates continued to have high case loads (~19 cases/day/100,000), while counties with mandates dropped to ~12 cases/day/100,000. That’s a pretty substantial difference and makes it clear that masks had an enormous protective effect.

Lyu and Wehby (2020) used a similar approach but at the state level (still within the USA). Once again, states with mask mandates enjoyed significant reductions in the spread of COVID relative to states without mandates. Indeed, they estimated that the mask mandates that were in place prevented 230,000–450,000 cases over a 22-day period. A similar study in Ontario Canada (Peng et al. 2024) found that in areas with mask mandates, between June and December 2020, mask mandates prevented 290,000 cases and 3,008 deaths, saving 610 million Canadian dollars. That’s a pretty substantial savings compared to the minuscule cost of masking (also see similar results for the state of Illinois from Castonguay et al. 2024).

I want to be clear here that I haven’t made any political assertions. I have simply shown the very large and robust body of data showing that wearing a mask helps prevent the spread of COVID. Mask mandates saved thousands of lives, and many more could have been saved if mandates had been universal. If you want to make the political argument that your freedom not to wear a mask is more important than saving those lives, you can do that, but you can’t claim that the mandates don’t work, because the data clearly shows that they do.

Fortunately, COVID numbers are low enough atm that masks probably aren’t necessary most of the time. However, flair ups and smaller outbreaks will continue to occur, and I’d encourage you to continue wearing a mask in those situations or any other time that you are at a high risk of exposure or spreading COVID (or other respiratory viruses for that matter).

COVID Vaccines

Let’s now turn our attention to the big one: vaccines. At the start, we need to clear up a few misconceptions. First, as explained here (see #10), the COVID vaccines were carefully tested prior to releasing them to the public. It is simply not true the vaccines were rushed or that scientists were pushing the vaccines before knowing if they were safe and effective. That said, now that billions of doses have been delivered, we have been able to conduct massive follow-up studies to further ensure that the original results were correct.

Second, many people seem to be under the impression that the vaccines didn’t work because they did not immediately, completely, and entirely stop the spread of COVID. I frequently see people post things like, “they lied when they said the vaccine would stop the spread of COVID.” In reality, no vaccine is 100% effective. The expectation was never that it would fully stop COVID in all cases. Rather, the expectations were that it would greatly reduce the spread of COVID and reduce the severity of infections when they did occur (both of which were correct). Additionally, the effectiveness of a vaccine at controlling the spread of a disease is always dependent on the percent of the population that receives the vaccine (i.e., herd immunity; see Coccia 2021). So, by refusing the vaccine, people were creating a self-fulfilling prophecy in which they reduced the overall effectiveness of the vaccine at preventing community transmission.

With all of that said, the COVID vaccines were still an enormous success and, as we’ll see, they have saved millions of lives.

As a caveat at the start, it is worth stating that many different COVID vaccines have been developed, and they vary somewhat in their effectiveness and safety; however, the overarching picture is that they are highly safe and effective (see this large comparison of vaccine brands: Toubasi et al. 2022), so in many cases, studies considered multiple vaccine brands together to look at the overall success of vaccination campaigns (usually pooled by vaccine type [e.g., mRNA]).

Let’s now look at the safety of these vaccines. All vaccines (and all medications) have a risk of side effects. This is nothing new to COVID vaccines. However, you have to remember that not vaccinating also has risks. In the case of COVID vaccines, studies have repeatedly found that most adverse reactions are mild, and serious complications are extremely rare. For example, in a large meta-analysis comparing multiple vaccine types, Kouhpaye and Ansari (2022) found that the most common adverse reactions were fever, fatigue, headache, pain, redness, and swelling and concluded that, “At the present moment the benefits of all types of vaccines approved by WHO, still outweigh the risks of them and vaccination if available is highly recommended.” Numerous other systematic reviews and meta-analysis have found similar results, with very few serious side effects (Amanzio et al. 2022; Haas et al. 2022; Bello et al. 2023). These results persist even if we look at meta-analyses on subsets of the population like the elderly  (Xu et al. 2023). Likewise, a meta-analysis on children aged 5-11 found that while there were minor side effects, there was “no increased risk of serious adverse events” (Piechotta et al. 2023).

Nevertheless, rare though they may be, serious side effects do occasionally occur. As with most vaccines, an allergic reaction is the most common serious side effect, but when looking across 15 studies with a total of 735,515 participants, Bello et al. (20223) only found 43 cases of anaphylaxis, none of which resulted in death. Indeed, in all the papers I have read, I have yet to encounter any reported deaths where the evidence compellingly showed that the COVID vaccine was the cause (see Lamptey 2021). It is a total myth that thousands of people died from the COVID vaccines.

What about myocarditis? This is the adverse effect that probably got the most attention. There is roughly a 2x increase in your risk of myocarditis following COVID vaccination (i.e., you are twice as likely to develop it; Juan Gao et al. 2023), but that is the relative risk (i.e.. how much your risk changes because of vaccination). To really understand the situation, we need to look at the absolute risk (i.e., how likely you actually are to develop myocarditis). A large increase in relative risk can still be a small increase in absolute risk. In the case of myocarditis, without vaccination (or COVID infection), the risk of myocarditis is 0.8–16.7 cases per 1 million people over a 30-day period. In other words, without COVID vaccination or infection, there is a 0.00008–0.00167% risk of someone developing myocarditis over a 30-day period. It’s a very small chance. Now, with the vaccine, that risk roughly doubles to 1.6–34.2 cases per million people, or 0.00016–0.00342%, which is still an incredibly small risk (numbers are from this meta-analysis: Alami et al. 2023). So yes, the risk increases, but your odds of developing myocarditis are still very, very low.

Additionally, we have to consider the fact that infection with COVID also increases your risk of myocarditis. Indeed, an actual COVID infection causes a much greater increase in risk than the vaccine causes. A massive study with millions of patients found excess rates (relative to background levels)  of 1–6 in the first 28 days after the first vaccine does and 10 in the first 28 days after the second vaccine dose, compared to 40 in the first 28 days following COVID infection (Patone et al. 2022). Likewise, another large meta-analysis found that a COVID infection was 7x more likely to cause myocarditis than was a COVID vaccine (Voleti et al. 2022). So, you have to look not only at the risk from the vaccine, but also the increased risk of COVID and myocarditis without the vaccine.

Now that we have looked at the risks, let’s turn our attention to the benefits, which are enormous. Meta-analysis after meta-analysis after meta-analysis shows that the COVID vaccines substantially reduce community transmission (i.e., reduce infection rates), reduce the severity of infections, reduce hospitalizations, and reduce deaths (Huang and Kuan 2022; Rahmani et al. 2022; Zheng et al. 2022; Wu et al. 2023).

Other scientists have gone beyond simply reporting the differences in rates between the vaccinated and unvaccinated and have used those numbers, along with population sizes, vaccination rates, and infection rates to calculate the number of lives saved by the COVID vaccines. Using this approach, Watson et al. (2022) estimated that from 8 December 2020 to 8 December 2021, across 185 countries, the COVID vaccines saved 14.4–19.8 million lives! So to anyone who thinks the vaccines were “rushed” or were being “forced onto the public” this is why we didn’t wait longer to release them and why scientists, doctors, and health agencies were campaigning so hard for people to vaccinate. No one was “weaponizing science.” Rather, scientists used the best evidence available to make a correct judgement call that saved millions of lives, and even more lives could have been saved if more people had taken the vaccine.

To be fair, not all analyses have come up with the same number of lives saved, but the number is still always in the millions. For example, Mesle et al. (2024) looked at lives saved only in European Union countries between December 2020 and March 2023 and concluded that COVID vaccines had saved 1.6 million lives in EU countries, which is a huge benefit, particularly when the risks associated with the vaccines are so small.

Finally, as scientists have been saying all along, the benefits aren’t just reduced infections and reduced deaths. Even when someone with the vaccine becomes infected, the infection tends to be milder. Several systematic reviews and meta-analyses have looked at this in the context of “long-COVID,” with the consistent finding that those who were vaccinated before getting a COVID infection are less likely to have persistent “long-COVID” symptoms compared to people who became infected without first receiving the vaccine (Notarte 2022; Watanabe et al. 2023).

Likewise, a huge meta-analysis of over 24 million people found the following (Ikeokwu et al. 2023; my emphasis):

Being unvaccinated had a significant association with severe clinical outcomes in patients infected with COVID-19. Unvaccinated individuals were 2.36 times more likely to be infected, with a 95% CI ranging from 1.13 to 4.94 (p = 0.02). Unvaccinated subjects with COVID-19 infection were 6.93 times more likely to be admitted to the ICU than their vaccinated counterparts, with a 95% CI ranging from 3.57 to 13.46 (p < 0.0001). The hospitalization rate was 3.37 higher among the unvaccinated compared to those vaccinated, with a 95% CI ranging from 1.92 to 5.93 (p < 0.0001). In addition, patients with COVID-19 infection who are unvaccinated were 6.44 times more likely to be mechanically ventilated than those vaccinated, with a 95% CI ranging from 3.13 to 13.23 (p < 0.0001). Overall, our study revealed that vaccination against COVID-19 disease is beneficial and effective in mitigating the spread of the infection and associated clinical outcomes.

In summary, the data clearly vindicate scientists and health officials and show that COVID vaccines were an enormous success. Releasing the vaccines saved millions of lives and prevented countless infections and hospitalizations. Further, the vaccines have proven themselves to be highly safe, and none of the doomsday predictions from conspiracy theorists and anti-vaccers have come true. With billions of doses given, it would be really obvious if these vaccines were actually dangerous. Hospitals would be filled with people dying from the vaccines. That isn’t happening, because the vaccines are very safe.

This continues to be relevant today. COVID has mercifully shifted to an endemic stage (thanks in part to vaccines), but it will continue to be a threat for the foreseeable future, and staying up to date with your COVID vaccination status continues to be a great way to protect yourself and those around you. Further, misinformation about the COVID vaccines seems to be eroding people’s trust in vaccines more generally. Earlier this week, for example, the USA had its first measles death in a decade. The unfortunate child was not vaccinated and was in an area with low vaccine coverage. Their death is a direct consequence of science denial.

Ivermectin and Hydroxychloroquine

Now that we’ve looked at two extremely successful interventions, let’s look at to two major failures: ivermectin and hydroxychloroquine. I’m going to keep this section nice and short: neither of them works at preventing or treating COVID. Both drugs received extensive testing (thanks to their promotion by certain politicians and political pundits), and they both failed that testing. Here, for example, are multiple meta-analyses and systematic reviews on ivermectin that failed to find a benefit (Deng et al. 2021a; Roman et al. 2021; Marcolino et al. 2022; Shafiee et al. 2022). Hydroxychloroquine is the same story, with numerous systematic reviews and meta-analyses finding that it is not effective at treating COVID (Kashour et al. 2020; Deng et al. 2021b; Mitja et al. 2022; Hong 2023; Lucchetta et al. 2023; Kaushik et al. 2024). It’s also worth mentioning that you can find studies looking at different doses, age groups, severity, etc. and they all tell the same story.

Further, these drugs are not without risks. Ivermectin is usually well tolerated with low side effects, but it still has side effects (Roman et al. 2021). Hydroxychloroquine appears worse, with far more studies reporting side effects (Izcovich et al. 2022; Hong 2023; Kaushik et al. 2024). To be clear, most of the side effects are fairly mild (e.g., nausea and diarrhea), but given a lack of benefit, the risk assessment clearly does not come out in its favor.

At this point, it is important to acknowledge that if you dig around, you can find some studies that disagree and do report benefits of either ivermectin or hydroxychloroquine, but there are multiple critical caveats, and it is important to understand how to assess the scientific literature. First, there were several really bad studies early on which were later retracted, but some reviews/meta-analyses included those studies, which massively biased the results. Second, even in the more reliable studies, the effect size is generally small, the statistics are weak, and the result is inconsistent. For example, Song et al. (2024) found that ivermectin did not reduce mortality or the time until a PCR negative test, but there was a slight reduction in the odds of someone needing to be on a ventilator. To explain what I mean by “slight” for these sorts of relative risk assessments, 1 = no effect, <1 = reduced risk, >1 = increased risk, and for something to be statistically significant, we usually want the 95% confidence intervals to exclude the number 1 (i.e., we are 95% confident that the true result is either less than or greater than 1). In Song et al., the effect for ventilation was 0.67 with a 95% confidence of 0.47–0.96. So it technically cleared the bar for statistical significance (i.e., confidence interval did not include 1), but it is a really weak result with the confidence interval almost containing 1, and it is a result that other studies didn’t support. Similarly, when Garcia-Albeniz et al. (2022) looked at the ability of hydroxychloroquine at preventing COVID, they technically found an effect, but it was very weak: 0.72 with a 0.55–0.95 confidence interval. As a final example, Ramdani and Bouazza’s 2023 meta-analysis of low dose hydroxychloroquine for treating COVID found a slightly significant benefit for reducing mortality (0.73, confidence interval = 0.55–0.97), but failed to find a reduced risk of needing a ventilator or being admitted to the ICU. Further, even, the significant reduction in mortality disappeared when they limited their study to only include randomized controlled trials (the most robust type of trial).

This pattern of most studies failing to find an effect with a few scattered papers finding inconsistent, small, barely significant results that disappear depending on how you look at the data is exactly what we expect for something that does not work. I’ve explained this before in more detail here, but for any well studied topic, there is going to be statistical noise. There are going to be papers that disagree and statistical outliers, but, when a treatment actually works, we expect a strong agreement that it works (a consistent body of evidence) with a large, reliable effect size, and a few dissenting papers that usually have small sample sizes, biases, or other methodological issues. In contrast, when something doesn’t work, we expect most studies to find no effect, with a mixture of small effect sizes and low significance, especially when looking at lower quality studies. Note that the expectation is never 100% agreement.

We can see these two predictions play out nicely in the topics of this post. Look at the difference in the evidence base for masks and vaccines compared to ivermectin and hydroxychloroquine. Sure, you can find a few studies that disagree on masks and vaccines, but the overarching effect is strong agreement among the highest quality studies that there is a substantial and consistent benefit. In contrast, on ivermectin and hydroxychloroquine, sure you can find a few weak studies that suggest they might have a benefit, but the overarching picture is that the majority of studies show that they don’t work, and the studies that say they do have such small effect sizes and/or marginal significance that they are not compelling. Stated another way, what this body of evidence shows is that if there is a benefit to ivermectin and hydroxychloroquine for treating COVID, it is very small and unreliable, which contrasts strongly with the large, reliable benefits we see for masks and vaccines.

Conclusions and significance

To sum all of this up, numerous large studies have vindicated scientist’s/doctor’s/health agencies’ decisions during COVID. Mask mandates and the COVID vaccines saved millions of lives, and refusing to implement them or delaying them would have cost lives. Many more lives could have been saved if more people had masked and been vaccinated. Conversely, ivermectin and hydroxychloroquine have failed to live up to their hype. It was wildly irresponsible of prominent politicians to promote them and claim, without evidence, that they were effective at dealing with COVID. I can’t help but wonder how many people needlessly died because they listened to politicians, not scientists, and as a result, relied on ivermectin or hydroxychloroquine instead of vaccines.

Beyond those deaths, just think about how much funding has been squandered studying ivermectin and hydroxychloroquine. There are hundreds of papers on them, easily having required millions if not billions of dollars and incalculable effort. How much better would we all be if that much effort and money had been invested in more promising treatments? Under normal circumstances, that is what would have happened. Scientists would have conducted some trials, found little or no evidence that they worked, and moved on. However, because it became a political issue with one side baselessly insisting that those treatments worked, scientists felt obligated to study them exhaustively and put way more effort into it than they would have otherwise.

The point is that, taken as a whole, we are clearly far better off relying on scientific studies than conspiracy theories, and listening to actual scientists and doctors than to politicians and pundits who think they know better than experts. Millions of lives were saved thanks to scientists, and countless more could have been saved if politicians backed scientists instead of attacking scientists/doctors and sowing fear, doubt, and conspiracy theories.

Related posts

Literature cited

  • Alami et al. 2023. Risk of myocarditis and pericarditis in mRNA COVID-19-vaccinated and unvaccinated populations: a systematic review and meta-analysis. BMJ Open 13:e065687
  • Amanzio et al. 2022. Adverse events of active and placebo groups in SARS-CoV-2 vaccine randomized trials: A systematic review. The Lancet Regional Health 12:100253
  • Baier et al. 2022. Effectiveness of Mask-Wearing on Respiratory Illness Transmission in Community Settings: A Rapid Review. Disaster Medicine and Public Health Preparedness 17:e96
  • Bello et al. 2023. Adverse Events Related to SARS-Cov-2 Vaccination: A Systematic Review and Meta-Analysis. Journal of Epidemiology and Public Health 8:284-297
  • Castonguay et al. 2024. Estimated public health impact of concurrent mask mandate and vaccinate‑or‑test requirement in Illinois, October to December 2021. BMC Public Health 24:1013
  • Chen et al. 2024. Universal masking during COVID-19 outbreaks in aged care settings: A systematic review and meta-analysis. Ageing Research Reviews 93”102138.
  • Coccia 2021. Optimal levels of vaccination to reduce COVID-19 infected individuals and deaths: A global analysis. Environ Res 204:112314
  • Deng et al. 2021 Efficacy and safety of ivermectin for the treatment of COVID-19: a systematic review and meta-analysis. QJM: An International Journal of Medicine 114:721–732
  • Deng et al. 2021b. Efficacy of chloroquine and hydroxychloroquine for the treatment of hospitalized COVID-19 patients: a meta-analysis. Future Virol.
  • Ford et al. 2021. Mask use in community settings in the context of COVID-19: A systematic review of ecological data. E Clinical Medicine 38:101024.
  • Garcia-Albeniz et al. 2022. Systematic review and meta-analysis of randomized trials of hydroxychloroquine for the prevention of COVID-19. Eur. J. Epidemiol. 37:789-796
  • Haas et al. 2022. Frequency of Adverse Events in the Placebo Arms of COVID-19 Vaccine Trials: A Systematic Review and Meta-analysis. JAMA Network Open 5:e2143955
  • Hajmohammadi et al. 2023. Effectiveness of Using Face Masks and Personal Protective Equipment to Reducing the Spread of COVID 19: A Systematic Review and Meta Analysis of Case–Control Studies. Advanced Biomedical Research 12:36
  • Hong 2023. Safety and efficacy of hydroxychloroquine as prophylactic against COVID-19 in healthcare workers: a meta-analysis of randomized clinical trials. BMJ Open 13:e065305
  • Huang and Kuan 2022. Vaccination to reduce severe COVID-19 and mortality in COVID-19 patients: a systematic review and meta-analysis. European Review for Medical and Pharmacological Sciences 26: 1770-1776.
  • Huang et al. 2022. The Effectiveness of Government Masking Mandates On COVID-19 County-Level Case Incidence Across The United States, 2020. Health Affairs 41: 445-453
  • Ikeokwu et al. 2023. A Meta-Analysis To Ascertain the Effectiveness of COVID-19 Vaccines on Clinical Outcomes in Patients With COVID-19 Infection in North America. Cureus 15:e41053
  • Izcovich et al. 2022. Adverse effects of remdesivir, hydroxychloroquine and lopinavir/ritonavir when used for COVID-19: systematic review and meta-analysis of randomised trials. BMJ Open 12:e048502
  • Juan Gao et al. 2023. A Systematic Review and Meta-analysis of the Association Between SARS-CoV-2 Vaccination and Myocarditis or Pericarditis. American Jurnal of Preventative Medicine 64:275-284.
  • Kashour et al. 2020. Efficacy of chloroquine or hydroxychloroquine in COVID-19 patients: a systematic review and meta-analysis. J Antimicrob Chemother
  • Kaushik et al. 2024. Effect of Hydroxychloroquine and Azithromycin Combination Use in COVID-19 Patients – An Umbrella Review. Indian Journal of Community Medicine 49:22-27.
  • Kouhpaye and Ansari 2022. Adverse events following COVID-19 vaccination: A systematic review and meta-analysis. Int. Immunopharmacol. 109:108906
  • Lamptey 2021. Post-vaccination COVID-19 deaths: a review of available evidence and recommendations for the global population. Clin. Exp. Vaccine Res. 10:254-275
  • Lu et al. 2023. Masking strategy to protect healthcare workers from COVID-19: An umbrella meta-analysis. Infection, Disease, and Health 28: 226-238.
  • Lucchetta et al. 2023. Hydroxychloroquine for Non-Hospitalized COVID-19 Patients: A Systematic Review and Meta-Analysis of Randomized Clinical Trials. Arq. Bras. Cardiol. 12:e20220380
  • Lyu and Wehby 2020. Community Use Of Face Masks And COVID-19: Evidence From A Natural Experiment Of State Mandates In The US. Health Affairs 39:1419-1425
  • Marcolino et al. 2022. Systematic review and meta‑analysis of ivermectin for treatment of COVID‑19: evidence beyond the hype. BMC Infectious Diseases 22:639
  • Mesle et al. 2024. Estimated number of lives directly saved by COVID-19 vaccination programmes in the WHO European Region from December, 2020, to March, 2023: a retrospective surveillance study. Lancet Respir. Med. 12:714-727
  • Mitja et al. 2022. Hydroxychloroquine for treatment of non-hospitalized adults with COVID-19: A meta-analysis of individual participant data of randomized trials. Clin Transi Sci 2023:524-535
  • Notarte 2022. Impact of COVID-19 vaccination on the risk of developing long-COVID and on existing long-COVID symptoms: A systematic review. eClinical Medicine 53: 101624
  • Peng et al. 2024. Impact of community mask mandates on SARS-CoV-2 transmission in Ontario after adjustment for differential testing by age and sex. PNAS Nexus 3:1-8
  • Piechotta et al. 2023. Safety and effectiveness of vaccines against COVID-19 in children aged 5–11 years: a systematic review and meta-analysis. Lancet Child Adolsc. Health 7:379-91.
  • Patone et al. 2022. Risks of myocarditis, pericarditis, and cardiac arrhythmias associated with COVID-19 vaccination or SARS-CoV-2 infection. Nature Medicine 28: 410-422
  • Rader et al. 2021. Mask-wearing and control of SARS-CoV-2 transmission in the USA: a cross-sectional study. The Lancet Digital Health3: e148-e157
  • Rahmani et al. 2022. The effectiveness of COVID-19 vaccines in reducing the incidence, hospitalization, and mortality from COVID-19: A systematic review and meta-analysis. Frontiers in Public Health
  • Ramdani and Bouazza’s 2023. Hydroxychloroquine and COVID-19 story: is the low-dose treatment the missing link? A comprehensive review and meta-analysis. Archives of Pharmacology 397:1181–1188.
  • Roman et al. 2021. Ivermectin for the Treatment of Coronavirus Disease 2019: A Systematic Review and Meta-analysis of Randomized Controlled Trials. Clinical Infectious Diseases
  • SeyedAhmad et al. 2023. The Effectiveness of Face Masks in Preventing COVID-19 Transmission: A Systematic Review. Infectious Disorders 23:19-29
  • Shafiee et al. 2022. Ivermectin under scrutiny: a systematic review and meta-analysis of efficacy and possible sources of controversies in COVID-19 patients. 19:102
  • Song et al. 2024. Ivermectin for treatment of COVID-19: A systematic review and meta-analysis. Cell 10:e27647
  • Toubasi et al. 2022. Efficacy and safety of COVID‐19 vaccines: A network meta‐analysis. J. Evid. Based Med. 14:245-262
  • Viera 2024. Effect of Face Mask on Lowering COVID-19 Incidence in School Settings: A Systematic Review. Journal of School Health 94: 878-888
  • Voleti et al. 2022. Myocarditis in SARS-CoV-2 infection vs. COVID-19 vaccination: A systematic review and meta-analysis. Frontiers Cardiovascular Medicine 9
  • Watanabe et al. 2023. Protective effect of COVID-19 vaccination against long COVID syndrome: A systematic review and meta-analysis. Vaccine 41:1783-1790.
  •  Watson et al. 2022.  Global impact of the first year of COVID-19 vaccination: a mathematical modelling study. Lancet 22:1293-1302
  • Wu et al. 2023. A systematic review and meta-analysis of the efficacy of N95 respirators and surgical masks for protection against COVID-19. Preventative Medicine Reports 36:102414
  • Wu et al. 2023. Long-term effectiveness of COVID-19 vaccines against infections, hospitalisations, and mortality in adults: findings from a rapid living systematic evidence synthesis and meta-analysis up to December, 2022. The Lancet Respir. Med. 11:439-452
  • Xu et al. 2023. A systematic review and meta-analysis of the effectiveness and safety of COVID-19 vaccination in older adults. Frontiers Immunology 14
  • Zeilinger et al. 2023. Effectiveness of cloth face masks to prevent viral spread: a meta-analysis. Journal of Public Health 46:e84-e90.
  • Zheng et al. 2022. Real-world effectiveness of COVID-19 vaccines: a literature review and meta-analysis. International Journal of Infectious Diseases 114:252-260.

Posted in Vaccines/Alternative Medicine | Tagged , , , , | Comments Off on Masks and COVID vaccines were huge successes; ivermectin and hydroxychloroquine were not

Incredulity fallacy: I’m right because I can’t believe I’m wrong

Posted on May 16, 2024 by Fallacy Man

incredulityI want to briefly discuss a logical fallacy that is surprisingly common, despite being so obviously absurd. I suspect that most people committing this fallacy do so without ever actually contemplating what they are saying, and it is my hope that discussing this fallacy will help people to both spot and avoid it.

The fallacy in question goes by many names including, argument from incredulity, personal incredulity, appeal to incredulity, appeal to personal incredulity, and argument from personal incredulity. I will simply refer to it as the incredulity fallacy.

“Incredulity” refers to an inability or unwillingness to consider the possibility of something, and the fallacy occurs when someone asserts that they are right about something because they cannot personally imagine that the alternative is true.

Stated like that, hopefully the problem is obvious. Reality is not determined by an individual’s imagination. Whether or not someone can imagine that something is true has absolutely zero bearing on whether or not it is true, and someone’s lack of imagination is not a good argument against something. Nevertheless, this argument is pervasive.

To illustrate this, I’m going to use an example that I have personally run into countless times over the past few years, and I suspect many of you have encountered it as well. The argument is from the political world, but I’m not making any political comments here. I’m just using this one particular argument to illustrate the fallacy.

I know many people, including family members, who think that Trump won the 2020 presidential election, and when I ask them why they don’t accept the results, one of the answers that I inevitably get is something along the lines of, “come on, you don’t really think that Biden got 81 million votes, do you? There’s no way 81 million people voted for Joe Biden!”

This is always stated as if it is a concrete, established fact that 81 million people could not possibly have voted for Biden. In reality, of course, it is nothing other than a self-reinforcing projection of their existing views.

We can rephrase it simply as, “I cannot personally imagine that 81 million people voted for Joe Biden, therefore it didn’t happen.”

The argument is, of course, entirely circular. We can rephrase it yet again as, “I don’t believe Joe Biden won because I don’t believe Joe Biden won.”

At no point are facts or evidence injected into the view. The view is simply stated as a fact.

While that example came from politics, this flawed reasoning also pervades anti-science arguments. For example, I have frequently encountered anti-vaccers who make statements like, “I can’t accept that injecting chemicals into an infant is good for them” or climate change deniers who say things like, “you honestly think that humans producing a little CO2 will alter the climate? lol.” These are both examples of incredulity. Just because someone doesn’t personal think “chemicals” can be good or human CO2 emissions can change the climate doesn’t mean those positions are false.

You’ll also notice that this fallacy is closely related to (and in some cases identical to) the appeal to common sense fallacy or just “gut feelings.” In all of these cases, the arguments are about what someone “feels” is true or false, not about what is actually verifiably true or false.

Going back to one of my interactions with a family member, when I pressed them on why they refused to accept that Biden won in 2020, they said, “It’s just common sense!” The problem, of course, is that “common sense” is not objective and varies from one person to another. As I pointed out to them, my “common sense” tells me that Biden won, so whose common sense should we listen to? If “common sense” is a reliable indicator of reality, then why do our “common senses” disagree with each other?

Likewise, I constantly encounter people who insist that it is just “common sense” that vaccines are dangerous, climate change isn’t being cause by humans, GMOs are dangerous, etc. Meanwhile, my “common sense” says the opposite. So, whose “common sense” is right?

This is, of course, precisely why we conduct rigorous scientific studies (and conduct elections) rather than just asking random people what they think. Sticking with the election example for a minute, if we follow these people’s arguments though to their logical conclusions, then there is no reason to hold an election. All we have to do is ask them who they think would win and we have our result. That, of course, is madness.

All of these brings me back to two central points that I make over and over again on this blog. Honestly, they are the two points at the very core of this blog’s existence.

First, fact check everything using good sources. Take nothing for granted. Verify that something is true before you believe it and verify that it is false before you reject it.

Second, be willing to be wrong. If you are someone who has made statements like the ones illustrated in this post, then really stop and think carefully about your views. Doesn’t it bother you that your argument boils down to “I’m right because I know I’m right?” Don’t you want to have a view of the world that is actually based on evidence and facts?

Always ask yourself, “what evidence would convince me that I am wrong?” If you say that nothing will convince you that you are wrong, then your position is, by definition, willfully ignorant.

Set aside your biases and preconceptions, accept the possibility that you might be wrong, and actually look at the evidence. Don’t trust your “gut instincts” or “common sense” or what you “just know.” Look at what is true, not what you want to believe is true.

Related posts

Posted in Rules of Logic | Tagged , | Comments Off on Incredulity fallacy: I’m right because I can’t believe I’m wrong