Goodhart's Law: When a measure becomes a target, it ceases to be a good measure.
this post was submitted on 15 May 2024
954 points (98.2% liked)
Microblog Memes
6021 readers
1417 users here now
A place to share screenshots of Microblog posts, whether from Mastodon, tumblr, ~~Twitter~~ X, KBin, Threads or elsewhere.
Created as an evolution of White People Twitter and other tweet-capture subreddits.
Rules:
- Please put at least one word relevant to the post in the post title.
- Be nice.
- No advertising, brand promotion or guerilla marketing.
- Posters are encouraged to link to the toot or tweet etc in the description of posts.
Related communities:
founded 1 year ago
MODERATORS
My state'a standardized test says kids are "at risk" if they aren't in the top 40% of the test. The top 50% could all be traditional "a-b" students. But because they weren't in the top 80% of a-b students they are at risk for failing academically.... It's so asinine and disheartening. The last half of the year is devoted to this idiotic test. Kids could be learning stuff that will enrich themselves... Instead they are learnig how to take a test better.
I was skeptical of these numbers until I did the research. It's even worse than I thought. Did you know that only 20% of students test in the top quintile? Is our children learning?
I was born extremely smart. I always tested way at the top of the curve... far superior to all my peers.
One could construct tests where this isn't a given. Create an exam with one really easy question and one really hard question. You'll end up with a huge spike in the center and two tiny tails. The top quintile of correct answers (<=1.6 correct answers) will be vanishingly small.
Do you statistics? This guy clearly statistics.
News at 11: "Half of people are below average!"
Yeah but the average is pretty fucking dumb.
Ashktually half people below the median.
Ackchyually, they never said which average they meant, you just assumed mean.
True, but those who know what "median" means probably also know what a "quartile" means, so if I used "median" it would've made my comment less of an "obvious, duh!" thing and spoil the unstated point I'm making as well as the joke.
Best leave the mathematical incorrectness there to preserve the feeling of obviousness.
Same thing in a normal distribution, no?
Consider an exam in which there are two questions: one very easy and one very hard. You'll get a supermajority of people who answer the first question and two tiny tails - zero correct, two correct - such that the mode is very high and the outlayer groups are very small.
Then well over half the people are in the median and mean.
Yes. In a normal, or Gaussian, distribution, the data is symmetrically distributed around the mean and thus mean (average value), mode (most frequent value) and median (middle value) all fall on the same point, which is the highest point of the curve.
Not in Lake Wobegon!!
Only all the children in Lake Wobegon are above average. It's balanced out by them all turning into idiots when they become adults.
I can almost see a facebook post along the lines of:
One doesn't have to be a rocket scientist to understand how terrible things are, when fully 25% of the population are in the bottom quartile.
25%!? That's nothing. Half of people are below average. That's twice of what you quoted!
Tide goes in, tide goes out. Can't explain that.
The gravitational pull of the moon.
But you're right, I cannot explain why gravity even exists.
Oh FSS please don't cede a smart introspection to that billowing buffoon
(nice followup though)
view more: next ›