3 ms·

That doesn't prove causation. There can be many underlying causes that explain the success of educated people just as well. For instance, as already mentioned,

That doesn't prove causation. There can be many underlying causes that explain the success of educated people just as well. For instance, as already mentioned, it might be that kids that are smart or work hard tend to finish a higher education. It could also be that kids who are born in wealthier families do both get access to better education and get more opportunities afterwards because of their better network. That last possibility does a much better job at explaining what the Bloomberg article is about: the different chances of success based on the color of your skin. Black students that graduate still have only 6.7% chance of becoming millionaires comparing to 37% for white people.

adventured 9y ago
You're context dropping at the end. The critical context for the causation in question, is that for all races, education stepping higher coincides to substantial increases in likelihood of wealth attainment. There are no exceptions that see a backwards regression (ie move up in education and down in wealth likelihood), which further boosts that it is very clear causation. Each step sees the average move considerably higher for all races.
- jcranmer 9y ago
  No. It's correlation that's not a statistical fluke, but it's not causation. There are other potential explanations that the data doesn't rule out, such as wealth being highly inherited (i.e., rich people are begat from rich people) and wealthy children being pushed into higher tiers of education than poor people. The existence of these other explanations show that the data is not necessarily indicative of causation.
- dbajakkchs 9y ago
  “Correlation is not causation” cannot be refuted by “but look how strong the correlation is!” Eg. the sun always goes down after the clock on my wall read 6 PM . We cannot conclude that the clock on my wall is causing the sun to go down.

adventured 9y ago

You're context dropping at the end. The critical context for the causation in question, is that for all races, education stepping higher coincides to substantial increases in likelihood of wealth attainment. There are no exceptions that see a backwards regression (ie move up in education and down in wealth likelihood), which further boosts that it is very clear causation. Each step sees the average move considerably higher for all races.

jcranmer 9y ago
No. It's correlation that's not a statistical fluke, but it's not causation. There are other potential explanations that the data doesn't rule out, such as wealth being highly inherited (i.e., rich people are begat from rich people) and wealthy children being pushed into higher tiers of education than poor people. The existence of these other explanations show that the data is not necessarily indicative of causation.
dbajakkchs 9y ago
“Correlation is not causation” cannot be refuted by “but look how strong the correlation is!” Eg. the sun always goes down after the clock on my wall read 6 PM . We cannot conclude that the clock on my wall is causing the sun to go down.