Well said! I find myself nodding along. I’ve some side comments

…An example of over valuing a single experiment…

A common word for this in stat circles is ‘anecdata’ (meaning interesting anecdotal info that might eventually be a good subject for an actual study

3) Probably the most common error committed by MEDIUM authors is to confuse correlation with causation.

Yep. One of my fav books in Jr HS was How to Lie with Statistics. (admittedly, I was a somewhat atypical student) One example in there was ‘proof’ that the salaries of Presbyterian ministers were dictated by the price of rum in Jamaica.

Consider that evolution has probably not worked overtime to ensure us good health into old age.

There’s been some interesting theorizing on that. It’s possible we have grandmas to thank for our unnaturally long lives. Argument goes like this: We are unusual in that females reproductive phase is much shorter than lifespan; there may be evolutionary pressure for longer life as grandchildren, assisted by a grandmother would have a greater chance of survival than those children without.

There are numerous ways to go wrong here but many examples involve projecting from animals to humans.

Not just Medium authors on this one: the vilification of cholesterol comes from studies on rabbits (which, as herbivores, don’t normally consume cholesterol at all — and which, indeed, proves to be lifespan shortening). Nutritional “science” is loaded with nonsense on stilts. Want actual science: look to animal nutritional studies, but bear in mind that we are not rabbits or mice.

Finally, there are good reporters here, too. I would particularly point to Shin Jie Yong as someone who actually respects the science, is clear about small sample sizes and preliminary results.