This sounds like really important work. There are too many unanswered questions in medicine to keep studying the same ones ad infinitum, without clearly defining how we will recognize our answers when we see them. Hopefully your work will extend to the observational research space as well (where redundancy seems particularly pernicious) (?)
My point is that neither are done to any standard of competence because medical âthought leadersâ decided it was unnecessary to learn the mathematical tools that other scientific fields take for granted.
I wouldnât agree with that - EBM was introduced by medical thought leaders who decided it was necessary for physicians to learn to use the quantitative tools of epidemiology and biostatistics for optimal decision making. For example Clinmetrics was written in 1987 by the founder of clinical epidemiology - Alvan R Feinstein around the time that EBM was born.
Explain this result then (from June 2022):
Findings: In this survey study of 215 physicians, most respondents (78.1%) estimated the probability of a medical outcome resulting from a 2-step sequence to be greater than the probability of at least 1 of the 2 component events, a result that was mathematically incoherent (ie, formally illogical and mathematically incorrect).
Statistics is quite hard and even trained statisticians often misinterpret p-values. This is not new. Physicians spend a lot of time learning biology, physiology, drug interactions, pathophysiology etc. This context is essential in every aspect of medicine, including clinical decision-making. Without it, words like p-values or Bayesian decision-making are empty.
Knowledge generation and interpretation is largely based on teamwork. Or at least it is certainly fun when that is so. This includes working with physicians, pharmacists, statisticians, epidemiologists bioinformaticians, biologists, computer scientists etc. We learn from each one.
And with enough such experience one begins to see that there are different types of statisticians in the same way there are different types of physicians. I would definitely not want a theoretical statistician with little applied experience to design, conduct, or interpret my clinical trials. For that I go to applied biostatisticians with specific skills. In the same way, while we should certainly listen carefully when an ID physician recommends an antibiotic for a hospitalized patient with cancer, it would not be as wise to let them choose the chemotherapy alone.
I will be the first to admit that statistics is hard, but the JAMA problems are not that hard.
What kind of statistics is going to be taught when the students canât do algebra?
The idea behind the proposal is twofold. First, algebra generates more student failure and attrition than almost anything else. (One of the guest speakers at Aspen said that his one piece of advice to any college president looking to improve graduation rates would be to fire the math department. We laughed, but he didnât seem to be kidding.) Second, in many fields, algebra is less useful than statistics.
I can only explain this result if you assess its internal validity and present this to usâŚ