Quantios is hiring: help reshape the college admission process using data | Hacker News Reader

Quantios is hiring: help reshape the college admission process using data

quantios.com

23 pointsfrisco15 comments

puts on ex-Ivy League admission officer robe

Boy, it's a good thing they've got someone with admissions experience on their team... http://quantios.com/index.php?page=company

Wait, wha? They don't?

How is this different from any other snake oil being sold to parents of middle class white kids (rightly) worried that junior's at a disadvantage?

"We use numbers!"

That's great, but at the end of the day, the admission offices don't. The decisions made at the elite schools are either obvious or highly random/subjective.

at the end of the day, the admission offices don't

Oh, they are very, very highly numeric places -- they just concentrate on the kind of numbers that have to be burned after calculation. ("We only admitted how many black students this year? STOP. Pick up the green folders pile and find me students you think would do well if they stretched a little bit.")

I used to like Japanese universities for doing admissions via high-stakes entrance exams. Having worked on those systems, I think I now love them. Publishing the score distribution for the exam last year, broken down by department, with the exact passing score needed for admission in the bloody application packet is just the cherry on top of the pie of awesome.

If Harvard leaked as much data as a Japanese university there would be a public firestorm. (Because Harvard engages in severe, systematic racial discrimination, but prefers to lie about the existence and extent of it.)

>highly random/subjective.

Part of the reason they seem so random is because we don't know their methods.

For example, most people don't know that many schools use email click-through statistics to gauge interest. Demonstrated interest affects your chances to get into a school (Cornell does this).

This is not a solution. This is part of the problem.

College admissions and my senior year was hell. I worked hard so that I could impress the Ivies which rejected me. I don't have any regrets about my academics but many of my applications were not worth the stress.

> This is not a solution. This is part of the problem.

I agree. I just wouldn't want to encourage this nightmarish process.

The process will be nightmarish whether or not you encourage it. It will continue to get even more nightmarish. There will never be enough space at these schools, and they don't want there to be enough space.

What Quantios is trying to do is build a system that will show students good matches for colleges that fit their profile. College admissions will always be annoying but perhaps it will save students some of the time the time and frustration you went through.

It seems that the service is to make the system _appear_ more rational, to put a lens in front of The System, offer a consulting service, and offer an advantage to students that are willing to play the game with parents that take the initiative.

Why is the college admissions process ruled by for profit organizations?

Looking back, I spent too much time worrying about my scores and not enough time making sure my essays reflected my unique achievements and skills.

  Looking back, I spent too much time worrying about my scores and not enough time making sure my essays reflected my unique achievements and skills.

That's still framed in the broken system in place now. Can your really tell someone enough about you in a 2-page essay for them to make an informed decision about your application? Test scores are just as bad, if not noisier, signals; they're just easy to grok numbers, so like one of the other posters mentioned, they come with implicit authority. That entire process is broken; we're not trying to sell another snake oil "what are my chances" calculator. There's real science behind what we're working on, and it's not a consulting service for the wealthy.

> We believe that the entire world is quantifiable. Do you?

In my experience, this belief causes many serious problems. I hope it's not considered rude to express this here on HN, but I would suggest considering possible ethical issues before taking a job like this.

Can you give me an example of an ethical issue arising from this view? Quantifying something doesn't impose a system of beliefs on its own.

> Quantifying something doesn't impose a system of beliefs on its own.

I suppose that is true, but in practice, quantifications are often used by a system that does impose certain beliefs. When this happens (and even when it doesn't) those beliefs are seen to carry an air of objectivity simply because there are now numbers behind them.

It is extremely tempting to treat subjective views as objective truth when there are fancy algorithms and numbers behind them. This leads to a willingness to impose certain views on people who disagree, and that's where the ethical issues can arise.

In the case of this particular product, there are related but separate issues having to do with encouraging the college admissions circus. Even if the intention is to improve it, the product would at best have the effect of propagating the myth that the college admissions process is fundamentally sound and can be perfected. In reality, adding more mathematical algorithms to the process will just give it a pretense of objectivity -- again, potentially harmful to millions of people.

My first sentiment is that this had already been done - and done pretty well - by other groups. Collegedata.com has a decent system and makes admissions predictions (though it's a front for a student loan company). Mychances.net appears to be very active in this field, too. It isn't clear to me exactly what criteria they use for predictions at any given school, but they have stated that they use gpa, sat, sat subject tests, class rank, ethnicity, state of origin, legacy status, and sports recruitment. In other words, other people seem to have done a pretty good job of extracting what they can out of the quantifiable data.

I hope it's not rude for me to ask what this new group offers that those two, plus collegeboard, unigo, and a handful of others, do not?

Not to be negative but I saw their pitch at Duke and they actually embellished the amount of users on thier facebook app and stole information from other teams in a pitch competition

I'm sorry? Can you be more specific? Given what we're working on, reputation is all we have (albeit as a function of how well our software works). We'd never do either of those things.

Reply on news.ycombinator.com