Statistics is one field in particular where it's easy for non-experts to reach completely false conclusions.

True with complex stuff, but we talk about covid-cases/day/region...that's really something everyone should interpret correctly.

Yeah; some sibling comments here say Covid stats are difficult to interpret, but cite phenomena (non random samples, false positives, conflation of population and sample statistics, etc) that should be covered by any reasonable introductory statistics course.

In the US, that’s first or second year undergrad territory. Overseas, it’s often covered in high school.

It’s particularly depressing that science reporters regularly get these things wrong.

I can’t imagine what (lack of) credentials are necessary to land such a job, but I often wonder if a basic understanding of statistics is a disqualifying credential.

If you mean https://news.ycombinator.com/item?id=25811860, the comment was correctly flagged because the flamebait in it broke the site guidelines badly.

I see zero evidence of that, and it's a serious violation of the site guidelines to post like that. Please see https://news.ycombinator.com/newsguidelines.html. As it says there, if you're worried about abuse, email [email protected] and we'll look at the data.

One of the most important lessons I've learned from poring over this data is that 99% of the things that internet users describe as "obvious" are not only obvious but false. A strange phenomenon, and one we all need to work on overcoming if we're to have real conversation.

COVID statistics can be very very difficult to interpret correctly, especially at the beginning when data was scarce.

For instance, most people agree it is best to exclude totally random testing sites from COVID numbers because even though the false positive rate is very low, at the beginning so was the incidence of COVID which means most of the positive tests would be false positive.

Including them or improperly dealing with that data would be considered base rate fallacy

https://en.m.wikipedia.org/wiki/Base_rate_fallacy

This is just one statistical phenomenon that has to be dealt with. It’s complicated and sometimes messy but it absolutely requires top minds

Proper interpretation of even the simplest data is easy to get wrong. For example, I routinely see smart, well-informed people describe the number of Covid deaths reported in a day as the number of people who died of Covid that day.

Lots of highly educated people are on opposite side that lockdowns are very unneeded and government is overplaying/lying about the severity including the one who founded most valuable space company.

>including the one who founded most valuable space company.

I wonder why would he say that ;)

Taking a stats class does not a data scientist make. COVID statistics can be very very difficult to interpret, especially at the beginning when data was scarce.

Edit to add example : For instance, most people agree it is best to exclude totally random testing sites from COVID numbers because even though the false positive rate is very low, at the beginning so was the incidence of COVID which means most of the positive tests would be false positive.

Including them or improperly dealing with that data would be considered base rate fallacy

https://en.m.wikipedia.org/wiki/Base_rate_fallacy

This is just one statistical phenomenon that has to be dealt with. It’s complicated and sometimes messy but it absolutely requires top minds

Posting like this and https://news.ycombinator.com/item?id=25811860 and https://news.ycombinator.com/item?id=25802483 is unacceptable and will get you banned on HN, regardless of how wrong someone else is or you feel they are.

Please review https://news.ycombinator.com/newsguidelines.html and stick to the rules from now on.