Medical innumeracy

One of the odd things about my life is that I often end up in a position to notice the innumeracy of people with far more education than I have. I can forgive people who have pursued an education in literature or the performing arts for innumeracy, but I can't quite bring myself to forgive medical doctors, or worse yet mathematicians. Perhaps it's just schadenfreude for my more accomplished colleagues, but watching these people fall on the sword of their own imagined knowledge fills me with many emotions at once. After all other emotions pass I am filled with sadness. It doesn't have to be this way.

I wish the anemic education medical doctors get in biostatistics was upgraded with a lab where they coded. Of course coding, or computer programming, is not a neccesary skill for doctors. But what many people don't seem to realize is that given the ease with which one can code for statistical analysis these days, coding is actually the short cut to seeing statistics in action. Unfortunately it's a short-cut many people are simply too arrogant to take. And not just physicians. Sadly I've seen even mathematicians go into all kinds of mental convolutions instead of simply checking data. One recently told me to stop coding in R, because he viewed it as a "poor man's MATLAB." I bit my tongue not to explain to him that actually, while Octave or Scilab could be described as a poor man's MATLAB, R might be better derided as a language so easy to program in even poets could probably grasp the basics. It's almost too easy...which makes me wonder why more people don't check data well.

I wish science and medicine would come to an equilibrium where people felt compelled to published their raw data. Unfortunately, I see every day the innumeracy of medical doctors would prevent this from even being useful. One recent statement I had to edit sticks with me. A doctor claimed something like the following (I am leaving out some identifying details like the disease itself):

There is a disease with 12 % prevalance. Ten percent of cases have a first degree relative with the disease. Therefore 90% of cases are sporadic.

Interestingly enough in the same article the doc wrote that people do not understand risk. Clearly.

There is actually more than one obvious logical fallacy in what the man wrote, but I tend to think both physicians and people in general over a certain age are a lost cause. There was a time when evidence based medicine was some trendy new idea- and that was perhaps only three decades ago. These days, given the advances in science and computing, I hope it whouldn't take more than a first university degree in almost anything to see why statements like the one I recently saw are incredibly wrong. Yet these kinds of leaps of logic pepper the thinking of people I work with all the time, everywhere.

On the one hand, everyone else's innumeracy means people like me can have a job forever, building and demystifying datasets seems like some kind of intellectual magic that so few people can do it, you can command a decently paid job even without great credentials. On the other hand, everyone else's innumeracy means many people who work with statistics and data science participate in all kinds of modern day numerical alchemy- and that is as kindly as I can put it- without anyone catching on. Even regulatory bodies, some of which can barely regulate simple algorithms anyway. The only hoep I have is for is a better future- someone get the kids some laptops and tell them how to download something like Rstudio. I hope they themselves can figure out the rest.