How We’re Fooled By Statistics

Is punishment or reward more effective as feedback? Do new medical treatments really work? What about streaks in sport? Without considering regression to the mean, we are prone to making significant errors.

Is punishment or reward more effective for helping people learn. A lot of people would say different incentives motivate different people, or in different circumstances, but in psychology there is a sizable body of evidence that in order to learn skills, positive feedback is more effective. This fining has been verified not just with humans, but also with other species.

It was strange then that after Daniel Kahneman discussed this research with Israeli fighter pilot instructors that he was met with resistance. They found the opposite was true: when they reprimanded a cadet for performing poorly, he invariably improved, but if they praised a cadet for an excellent performance, the next attempt was not as good. In order to solve this apparent contradiction we first need to understand regression to the mean.

[seifer]

One of your better videos
It is unfortunate that your next likely will not be as good.

[littlebrother]

The book "the talent code" talks about the appropriate style of feedback according to skill level. A beginner needs more encouragement to improve, regardless of immediate outcomes, but relative expert is engaged and needs no encouragement, but they still require detailed feedback in order to correct their mistakes. Reversing those, I could see beginners giving up out of frustration, and experts rolling their eyes, feeling like their time is being wasted.

[Krekkertje]

Can we all take a moment to appreciate how good Derek is at drawing bell curves?

[MilanVucic]

I read about this exact example in Daniel Kahneman's book "Thinking, fast and slow". It is an amazing example of a fallacy we're primed to believe since a young age, which actually proves detrimental to both the performance and mental well-being of a person always being reprimanded. It is just one of the many quirks of our mind that Daniel explores in the book, a big recommendation.

[besmart]

So, theoretically, we could watch or not watch this video, like it or dislike it, share it or not share it, and the next Veritasium video will either be better or worse? :)

[Alex-fumd]

2:13 Question #1 was really, really, really, really unfair.

[misery]

I started a new job and my confidence in my skills isnt that high yet, but the guy teaching me believs in me. He sends me off doing stuff as if its nothing, even though it makes me nervous. That makes me proud, and I want to prove myself. So I try very very hard until I get it right. If he wouldnt believe in me, I'd be discouraged.

[shawnyellowbird]

This is ALOT of effort to explain our last Golf outing, dude...

[whippycream]

I appreciate that you had to film yourself 49 times for those classroom shots!

[BhanuPChauhan]

2:22
Name: Jimmy Neutron
Student # : 1.675 x 10^(-27)

[jerryoconnor]

I feel like regression of the mean is something I’ve applied so much in my life but never actually knew about.

[mattskord]

“The negative feedback was a harsh reprimand and the positive feedback was a 👍🏻”

Seems uneven but ok

[amaldabe]

I wish my teacher would say "Great work!" whenever I got a 61/100.

[WilliamBradey]

I wish my dad was smart enough to know this when I was growing up.

[ubaft]

Selecting both answers on the question "This statement is a lie" and receiving 1/2 points is next level IQ

[rsedan]

A great video. I kept wishing that you would clarify that it isn't that *the next* event is likely to be better (after doing unusually poorly) or worse (after doing unusually well). It is that *any* event is likely to be average (by definition). There is nothing special about the next event. Thanks for a thoughtful video!

[Yerg]

That distribution at 2:18 was drawn brilliantly

[Wifi_Cable]

This is why I love science. I clicked on this video thinking... why is a physicist talking about psychology? Why?
Because science has to take into account all sides, this is one many people probably overlook.

[lucyfyrearchoftwilight]

I'm surprised he didn't bring up the fact that the test was poorly executed anyway. What they should have done was to give negative feedback to both groups and then, to other test groups, give positive feedback to both groups. That would help reduce "regression to the mean" as a factor.

[BasenjiAdventures]

That regression needs an attitude adjustment. It’s so mean.