Hacker News new | ask | show | jobs
by throwaway9980 3063 days ago
Imagine that you could aggregate all the anecdotes shared on the internet. It would be fascinating to compare the statistical conclusions you could draw against accepted science.
2 comments

https://nequalsmany.com/

Check it out, this is happening. They are doing many crowdsourced analysis, the most prominent being a 30-day carnivore diet protocol: eat meat, drink water.

I’d be lying if I said I didn’t expect HN to tell me that it’s already being done. I am impressed and amused that it happened this fast and it’s pretty much exactly what I posited. Slow clap.
It would be damn near impossible to conclude anything from such research without a control. “Randomized double-blind, placebo-controlled trial” is the critical sentence - at best you’d get a correlation which probably has been influenced by the availability of research. Finding truth is freaking hard.
Do you think that qualitative data is useless?

'It would be damn near impossible to conclude anything from such research without a control. “Randomized double-blind, placebo-controlled trial” is the critical sentence'

Since the advent of randomized double-blind, placebo-controlled trial has our public health improved or declined?

I'm resisting the urge to respond flippantly, since I think that's a question that could be answered rigorously and is certainly worth asking.

With that said, since the advent of the RCT, life expectancy has gone up dramatically in the Western world. It would be hard to imagine testing the efficacy of, say, antibiotics or vaccines without testing them against a control in a placebo-controlled, double-blinded way.

Same goes for simple things like washing your hands and administering prophylactic antibiotics during a C-Section, which dramatically improves the odds that a mother will live through complications during child birth. There would be no way to actually know that these were important without studying them vs. the status quo at the time.