Hacker News Reader

Comment by the_jeremy | Hacker News Reader

It's still a huge improvement (not that this is good, just that a single metric for adversity is completely ridiculous).

View on HN

Some would argue that any metric for adversity is inherently unfair. Probably the best example to support this view is that of a student with hidden adversity, e.g. abusive parent or unreported learning disability that had to be overcome, despite being an "advantaged" ethnicity and living in an affluent suburb.

Agreed. I really don't believe the College Board can gain an accurate picture of whatever adversity their test takers have faced based off of the data they collect.

Of course not. It's always going to be shot in the dark, they'll have to put a ton of trust in the information put in front of them.

They aren't going out in the field to interview friends and family.

A single nice family home in the suburbs could have one kid turn out to be a successful judge and the other a drug addict drop-out (even Trump's older brother died an alcoholic). Good luck measuring that honestly from a college entrance office.

Do you believe that the consumers of the "landscape" scores won't sum those values up or apply some substantially similar reduction?

> a single metric for adversity is completely ridiculous

What about a single metric for scholastic aptitude, though? :)

Colleges probably place too much weight on, say, the SAT (which is, I presume, that to which you are referring), but they still use more. Years of report cards, letters of recommendation, extracurriculars, essays, AP exams, SAT subject tests, ACT, personal interviews, "soft factors", and other things that aren't disclosed. I agree it's certainly not perfect, and think it needs improvement, but personally can't think of a way to easily filter the huge numbers of applicants without quantifiable metrics.

The goal is to perform action selection. No matter how many metrics you come up with, you have to aggregate them into a single score to rank the set of possible actions. The fact that producing such a score is preposterous is merely an indication the whole idea of adversity metric[s] for deciding the future of our children is morally and functionally bereft.

The change moves the onus of metric aggregation from the College Board to individual colleges. We'll institutionalize injustice / poor decision making, but you'll have a choice of which institution will apply the injustice.

Blind admissions. We may even incentivize kids to actually study for college, instead of waltzing in with a seventh grade reading level. If we don't like that recent immigrants outperform natives [of all colors!], perhaps we could perform a serious root cause analysis and address it, instead of papering over the metrics.

https://www.campusreform.org/?ID=6174: The average college freshman reads at 7th grade level

Reply on news.ycombinator.com