pull down to refresh

The Statistical Significance Scamtowardsdatascience.com/the-statistical-significance-scam-db904be36714
278 sats \ 4 comments \ @south_korea_ln 12 Nov science
For once, a pretty good article on medium. A nice primer to how to and how not to interpret statistical results.
An example of this mistake happening time and time again is when there is a (statistically significant) discovery in carcinogens i.e. something that causes cancer. A 2015 Guardian article said:
“Bacon, ham and sausages rank alongside cigarettes as a major cause of cancer, the World Health Organisation has said, placing cured and processed meats in the same category as asbestos, alcohol, arsenic and tobacco.”
This is straight up misinformation. Indeed, bacon, ham and sausages are in the same category as asbestos, alcohol, arsenic and tobacco. However, the categories do not denote the scale of the effect of the carcinogens, rather, how confident the World Health Organisation is that these items are carcinogens i.e. statistical significance.
write
preview
related posts
152 sats \ 3 replies \ @SimpleStacker 12 Nov
For sure, many people, even in the science community do not properly understand statistics, p-values, or hypothesis testing.
Selection into publication is also a huge problem. If 100 studies are run on 100 random samples, all testing the same hypothesis , I think you should expect to get 5% of the studies with a p-value <0.05. If only studies with p<0.05 get published, then all the published studies will show a statistically significant relationship despite the 95 studies that didn't and weren't published.
(I think my above description is accurate, but I'd have to do some more review of my statistics foundations to confirm.)
reply
21 sats \ 1 reply \ @satgoob 12 Nov
There have been controversies recently around people "P-hacking their studies:
For reference: "P-hacking is a term used to describe various techniques that researchers can use to increase the chances of finding statistically significant results in their study, even if the results are not actually meaningful." https://www.physiotutors.com/wiki/p-hacking/
reply
3 sats \ 0 replies \ @SimpleStacker 12 Nov
Yeah, it's definitely an issue. Which is why people should not follow scientific studies uncritically. Science is useful but you have to understand the conditions under which it is produced.
reply
0 sats \ 0 replies \ @SimpleStacker 12 Nov
Slight addendum to my above post:
If the null hypothesis is true, you should expect 5% of the studies to have p<0.05.
reply