Michael S. Scheeringa :
What if there is another reason to explain the misleading information about a psychological pandemic due to COVID-19 besides the usual cynical explanations for why journalists and researchers publish hype? Perhaps there is a more innocent explanation – something deeper about a fundamental inability to understand what they are writing about. Specifically, are there some people who honestly have difficulty grasping science?
Many people have fundamental difficulties with math; just ask any high school algebra teacher. Many people are musically challenged and can’t play instruments. Watch any elementary school recess and you’ll see those who are coordination-challenged. Many people are directionally-challenged and get lost driving in familiar neighborhoods. Perhaps these journalists, researchers, and editors who write misleading stories are trying their best, but just struggle with the concepts of science. At the extreme end of the spectrum, we might even call it a science learning disability.
If you’ve been reading my series of blogs on mental health research during the COVID-19 pandemic, you know the drill. Researchers and journalists hyped misleading claims of a so-called massive, crushing, and overwhelming pandemic of psychiatric disorders that simply does not exist. The quality of research has been poor and misleading in large numbers. Researchers and editors are both to blame for this “garbage in, garbage out” hype about research for the past year.
It’s too easy to blame this on the usual suspects for hype. The usual suspects for researchers publishing garbage are the pressure to “publish or perish,” and the requirement to justify their importance in the science world. The usual suspects for journalists are the “if it bleeds, it leads” strategy to capture an audience, and the political bias that is often part of the media business model.
On March 2, 2021, a nonprofit organization that aggregates and analyzes health care data called FAIR Health, released a white paper which implied that mental health problems for children during the pandemic had increased massively. This was an unusual study because of the unorthodox way they reported their data. To take one example, the FAIR Health white paper reported that mental health insurance claims for teens as a percentage of all medical insurance claims increased 97.0 percent from the start of the pandemic in March 2019 to March 2020. This was reported immediately in many news outlets as mental health problems had skyrocketed due to the COVID-19 pandemic.
If you didn’t catch it already, go back and re-read what FAIR Health reported. It does not actually say that mental health claims increased at all. It simply said that mental health claims increased as a percentage of all medical claims.
I dug into the data and looked at their report and a previous report to get the base rate. Their own data actually show that the raw number of mental health claims actually DECREASED from March 2019 to March 2020. The short explanation is that the raw number of all medical claims dramatically decreased from March 2019 to March 2020. So the raw number of mental health claims actually decreased too, even though they increased as a percentage of overall medical claims. I estimate that in raw numbers the number of mental health claims decreased by about 1 million per month. Not a trivial drop.
I sent my longer explanation to FAIR Health and, to their credit, they agreed with me. Why FAIR Health opted to represent the findings the way they did in such a complicated and alarming fashion is another concern. My point is, when I first saw the FAIR Health claim on the TV news, I recognized in about one second that this was not saying that mental health claims for teens increased in absolute numbers. I might have what psychologists would call good “number sense.” I believe, however, that when many other people saw this who do not have good number sense, they believed this meant that mental health problems for teens had increased. In fact, that is how many in the media reported the study.
The New York Post reported “Those seeking help with mental health rose during that time, despite other medical claims declining” (Steinbuch, 2021). The University of Minnesota’s Center for Infectious Disease Research and Policy reported, “Mental health insurance claims for US teens roughly doubled early in the COVID-19 pandemic over the same period in 2019, according to a Fair Health report released yesterday” (Van Beusekom, 2021). Fox News reported “American teens seeking mental health respite nearly doubled insurance claims amid the coronavirus pandemic in 2020” (Rivas, 2021). Kaiser Health News reported “Teenagers’ Demand For Mental Health Care Skyrocketed During Pandemic” (Kaiser Health News, 2021). Those statements are not what FAIR Health reported and sounded a false alarm.
I’ve talked to a psychologist who conducts psychoeducational testing on children, and there is clearly no official science learning disability category. The traditional learning disabilities include math, reading, and written expression.
I believe there is something akin to a science learning disability in many individuals. Experience shows that many individuals are ill-equipped to understand science and statistics, and it can’t be remediated easily with education. For journalists, why would we assume that individuals who are good at writing are equally good at understanding science? We certainly don’t expect the reverse. For doctors, why would we believe that individuals who are good at interacting with humans for diagnosing and treating medical conditions are equally good at research? We certainly don’t expect the reverse. You don’t need to be a psychologist who conducts psychoeducational testing to know those are completely different skill sets.
I can only speculate then what a science learning disability might involve. One important skill seems to be the ability to understand cause and effect, which involves the ability to think in abstract terms of change in a construct over two time points. Another skill involves the ability to simultaneously consider opposing ideas, for example, that two constructs can act on each other but can also not act on each other (i.e., correlation might equal causation but correlation might not equal causation). Also, the capacity to imagine the unseen, such as the action of micro-organisms and chemicals that can’t be seen with the naked eye, seems important, which, strangely, also sounds like faith in the unseen (i.e., a virus can sneak through the side of a mask but a virus is also subject to gravity and can’t levitate to a person six feet away).
It might involve the mental braking power to suspend the more energy-efficient ideological belief systems and switch to the more energy-inefficient use of reason; this might even be more of a temperament trait than a cognitive skill. Having a good number sense is surely helpful. In addition, an entire body of psychological research exists to describe why humans are in general poor at understanding mathematical probabilities. In order to understand probabilities, it is usually necessary to slow down and consider base rates of how frequently something occurs at other points in time, which was the problem with the FAIR Health study; they didn’t report the base rates of medical claims.
If a business behaved this way, continuing to advise clients on topics which they themselves barely understood, the business would be facing lawsuits or bankruptcy. But the media and science journals stay the course, still insisting that they know how to understand science. The burden of proof that they know how to understand science should rest on those doing the publishing. Instead, what they’ve proven so far is that it is easy to misunderstand science.
(Michael S. Scheeringa, M.D., is Professor and Vice Chair of Research for psychiatry at Tulane University School of Medicine).
