Without introspection, without skepticism, without ethics, without respect for opposing viewpoints, an expert armed with statistics is, to borrow a line from Better Call Saul, “like a chimp with a machine gun.”
Without introspection, without skepticism, without ethics, without respect for opposing viewpoints, an expert armed with statistics is, to borrow a line from Better Call Saul, “like a chimp with a machine gun.”
X is an unknown property. Spurt is a drip under pressure. So an expert is an unknown drip under pressure.
That vid of the chimp with the ak is hilarious.. It’s what is needed when those gangs descend on stores and strip them bare.. as an aside tho.. with turdlala’s passing the bill to the senate, censoring communications, we will be forced back into the t-shirt, bumper sticker era, but what it won’t force us into is the watching of pm t’s subsidized gov.speak. It will infact speed up the death of those outlets when even the sleepyist amongst us realize that what they are seeing is just pravda. so come on folks lets get too it. piss all over these outlets with our comments,slurs,shunning,shamming, what ever we can come up with..make them look like the fools they are.. Have fun people, be careful out there.
As I recall – and it is really back in the day, a Nobel laureate identified an expert thusly:
ex – former
spert – a squirt under pressure.
I rest my case.
Expert : A guy from outa town who gets $800 Bucks a day plus expenses and tells you what yo have all ready tried to tell the boss.
Chimp with an AK-47
https://www.youtube.com/watch?v=QxYmm5yCJBg
This is even scarier. The chimps are passing down to their young the knowledge of how to use Machetes.
https://www.youtube.com/watch?v=d-0vbvy2ip4
This is an enjoyable read.
and…
https://iowahawk.typepad.com
It was a sad day when David gave up his site.
Paul Krugman is an idiot, and always has been. No statistics required to validate that.
https://news.yahoo.com/paul-krugman-always-wrong-never-173530058.html
A story never to be seen again.
Sally Clark’s sad story reminded me of when Shaken Baby Syndrome was all the rage in the early 2000’s. Once again…”expert testimony” sent mothers, fathers and caregivers to prison based on an absurd 60’s era medical experiment where chimps were strapped in sleds and smacked into a wall at 40 mph.
Might, however, explain some public health official’s intellect.
Interesting read. I read a lot on statistics and statistics misuse, and have encountered all these examples before.
Many (all?) journalists don’t understand statistics. And I suspect lawyers are the same.
Two other areas where even experts get fooled: randomness effects and using the normal distribution (so called Bell curve) to make forecasts. Economists are notorious for using normal distributions.
Nicholas Nassim Taleb does a nice job of exposing normal distribution misuse. Although his books could use some editing.
Math is hard. You should ring Taleb up. Not that I disagree. His prose isn’t quite summer reading for the beach.
It is common practice to improperly obtain population samples for statistics these days.
For starters, most population samples are too way too small (e.g., 1500 Canadians sampled) can not be generalized to the larger population (38,000,000 plus Canadians) such as to find out what “we Canadians “think or do”, thanks to diversity etc.
All statistical-generating firms, government and amateur sites are guilty, as most employees grew up with calculators and never actually calculated a standard deviation by hand. After all COVID was run by models, not facts.
The generalizations are even weaker when the stats report sub-populations of ages and sex, rural/urban etc., as the subgroups are tiny, like 750 males or 350 in the 30-45 age group or 70 rural. Logic 101; statistics 101!
Then, we have sampling recruitment issues. How did the company/university/newspaper find the people and communicate with them? There are still quite a few, increasingly lots of, Canadians who 1) do not speak English or French very well; 2) do not use computers much; 3) and/or would not fill out a phone, mailed or computer survey; 4) do not have a land phone with a published number; and 5) do not use a cell phone much; or 6) do not answer it. Good luck with that!
Then, we have irresponsible reporting: no news reports any methodology whatsoever, but statistics 101 is rarely taught.
Some of what stats can does is acceptable, but some of it can’t really be supported (usually the regional results that use a much smaller sample size than the population would expect).
When it comes to polling companies, as their job isn’t to gauge opinion but to influence it, first with push polls that have terrible questions, and then with the results that get published without commentary or methodology, often without a link to the poll in question. After all they can charge more for getting the results that you want.
Exactly
Stats Can used to hire very highly qualified stats people, and likely still does (I am out of the know, now). They should know better about sampling, but it has become more difficult, plus they have budgets. So they ignore ethics/professionalism/education, to keep the job…
You are right, JD, about other outlets who are driven by marketing managers, and statistians who get paid to produce what the outlet wants for preferred narratives.