Hacker News new | ask | show | jobs
by jimmyvalmer 1610 days ago
Basic understanding of statistics informs us that aggregate results do not predict individual outcomes.

Actually, that's precisely what it does.

2 comments

I'm not sure what you're arguing but it seems like some form of ecological fallacy.

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

I'm not even sure what statistics has to do with the OP's question or my responses. It only came up because you implied I don't understand "Bayesian likelihood" in your ad hominem.

My point is that a student can get a poor math grade or struggle in high school for reasons other than lack of ability or interest in math. One way to test that is to try learning math in a different environment (the personal example I gave). If a student is failing math one hypothesis is they are not able to understand math, or aren't applying themselves. Another hypothesis is that the teaching environment is less than ideal -- something Bayesian likelihood might support.

I'm not even sure what statistics has to do with the OP's question

Well, that was obvious. What's more likely? That OP is failing math because it's beneath him, or because he's got more in common with the *prior* million disaffected dropouts?

I see. I don’t really care about or think it’s relevant why the OP is struggling with math. My comments were suggesting alternatives to high school math classes that may have better outcomes. I took the OP at their word that they want to learn high school level math, which is I think obviously a different goal than getting a good grade in math class.
What? No it doesn't.

My life expectancy is 85. Does that mean I'm going to die at 85? Probably not. It means that, of the people of my age, gender, and race, the current statistically-based projections are that half of them will die by 85, and half won't. How does that predict the outcome of any individual?

We both know this argument wasn't referring to point estimates, but I am happy to engage in the pedantry of oneupsmanship. That's a big part of HN.
No, we don't both know that. If you didn't mean that, you worded your comment badly, because what I took from it is exactly what I responded to. (And, no thanks for assuming that I was responding to something I knew wasn't your point. That's against the site guidelines of charitable assumptions.)

Would you care to try stating your argument again?