Hacker News Reader

Comment by robocat | Hacker News Reader

> Harvard, for example, quadruples the patient’s income.

Bullshit. 98% of Harvard students graduate: https://college.harvard.edu/resources/faq/what-harvards-grad...

Questionable essay if it can’t get basic premises correct.

Edit: Selection bias. Rich people go to Harvard. Hardly surprising that Harvard graduates have different outcomes from the general population. For example:

  The median family income for Harvard undergraduates is $168,800—more than three times the national median, according to a recent study.

Edit: Don’t use Harvard as an example of university causing positive outcomes because the group of people that enter Harvard are not representative

View on HN

Your comment makes no sense to me. The OP links to https://www.collegesimply.com/colleges/massachusetts/harvard... , which states:

> 10 years after enrolling, the average income of former Harvard University students who are working and no longer in school is $136,700, which is 299% higher than the national median. Sources, U.S. Department of Education College Scorecard / Department of Treasury.

An increase of 299% means multiplying by (1 + 2.99) = 3.99x, which the OP rounds to 4x.

I dropped out of college and my earnings last year (still in my mid/late 20s, have easily another promo or job hop before 32) are over double the Harvard median, therefore dropping out of school causes you to earn 8x the national median.

"Doesn't count" you say? "Confounding variables" you say? "Causation and correlation aren't the same" you say? I'll just have to take your word for it, grad, it's not like I have a college degree after all!

P.S. I love rejecting Ivy League grads in interview loops. They've very consistently been the absolute worst performers on LC easies.

> P.S. I love rejecting Ivy League grads

Saying 'I love' makes you lose credibility and deprecates the points you are making.

> I dropped out of college and my earnings last year (still in my mid/late 20s, have easily another promo or job hop before 32) are over double the Harvard median, therefore dropping out of school causes you to earn 8x the national median.

Good luck going forward. This is a small slice of your life. Things can and do change with earnings. As many old timers might tell you there are always many 'changes of fortune' where people who are beneath you early on end up surpassing you later in life.

Biased much? Lording acceptance/rejection over any candidate seems like a trip

But to stay on topic, I generally agree with the consensus of this thread and your comment. The inputs and outputs here probably aren’t linked to the schooling itself.

I wouldn't say I lord it over anyone. I'm very polite to all candidates and make my decisions strictly based on technical performance.

That Ivy League grads tend to bizarrely be reliably poor in this regard (a surprising number fail to actually produce working solutions) simply amuses me - no amount of intergenerational wealth is a substitute for competence. Technical aptitude tests are the great equalizer between the haves and the have-nots.

That's not to say I've never given an Ivy League grad a thumbs up in the loop, but their performance tends to subvert popular opinion and expectations of "Ivy League grads" as a group.

In case anyone's curious, the group that most reliably aces the challenges I provide are first generation Chinese immigrants. It's basically the norm that they will not only produce working solutions to every problem, but that they complete their solutions quicker than other candidates, and that their solutions are computationally quicker and more efficient than those of other candidates as well.

I am not Chinese or an immigrant. I have no loyalty or allegiance to China. I would describe China's political system as apalling. Hypothetically, as a business owner with valuable IP, I'd be weary of first generation Chinese immigrants on espionage grounds. First generation Chinese immigrants might as well be the least likely group for me to offer preferential bias to.

That said, I can't argue with results. If you've got the skills, I couldn't care less about the reputation of the place you came from, regardless of whether it's a specific country, specific school, or lack thereof. Also, I'm not given enough of a vested interest in my employer's company to care about whether they suffer IP theft, nor is my job to prevent that, only to judge who among our candidates has the requisite aptitude and who doesn't.

I'm sure you are super-non-biased and a completely rational and logical being that has no grudge whatsoever against anyone or any institution

I think you're missing a /s

It works even better if you drop out of Harvard like Gates and Zuckerberg did!

> Your comment makes no sense to me.

That's because you didn't go to Harvard (/s) and you need to learn yourself what Selection Bias is.

If Harvard selected its students randomly from the general population of all 18-year-olds (including those who dropped out of high school), then it would be statistically relevant that Harvard grads make 299% more than the national median.

But this is not the case. Harvard chooses its students from the brightest, highest achieving, most highly motivated kids in the nation (and whose families tend to be wealthy.) It is reasonable to think that had Harvard vanished and these kids never got a college education, they would still succeed greatly in life.

So to mitigate the selection bias, we might compare Harvard grads versus the population of kids were in the top 5% of their high school class (but even that's not perfect.) The measured boost of the "Harvard Super Drug" would be considerably lower than 3x--maybe 10% if that.

FYI, robocat added his note about selection bias as an edit, AFTER I had posted my response.

With the new edits, robocat's comment now make sense to me.

That said, there's still a big caveat: We should look at the change in income for students drawn from the low-income population.

And please follow the guidelines: There's no need to lob insults.

> We should look at the change in income for students drawn from the low-income population

You'd really have to compare outcomes among a population of low-income students who equally "could have gone to Harvard" but some of them didn't for whatever reason. And even here there's a hazard of selection bias--as one example, a "reason" for not going to Harvard might be teenage pregnancy, which is certainly going to put a dent in economic outcome.

If you compare low income students in the top 5% of their HS class who went to state school vs. Harvard I'll bet the gap is smaller still. For example, here's a list of where Fortune 500 CEOs went to college--it's mostly state schools. Warren Buffett went to U. of Nebraska, Tim Cook to Auburn, Larry Page to U. Michigan, etc. https://twitter.com/daveckang/status/1667946905966637061

I had to add Selection bias afterwards, because I had thought it was bleeding obvious but user li_am made a snarky remark so I made the comment clearer.: I just don’t understand how you and li_am could totally miss the point, The 98% pass rate shows it so clearly. Note: I am not from the USA but I know this shit.

Their $33k seems suspiciously like half of the wonderful meaningless household income.

$42K seems to be more accurate unless they are getting to $32K by somehow correcting for age.

https://www.census.gov/library/publications/2021/demo/p60-27...

More opinion on why Harvard in particular is a bad example to use: laundering priviledge:

https://astralcodexten.substack.com/p/why-match-school-and-s...

Questionable comment if you can't get the basic premise of the article correct.

Reply on news.ycombinator.com