How would you reliably capture race, or sexual orientation, or gender identification — or even income — on a self-reported metric with no third-party validation? They aren't talking about sending the College Board your tax returns.
Realistically, this will be based on ZIP code and registered high school. Information that's already publicly-available. This changes nothing except potentially making neighborhood-level affirmative action more politically palatable.
I do believe with just a few pieces of the information they're taking into consideration (namely, neighborhood, if you have a single parent, and if you're ESL [not sure if they're talking about the child being ESL or the parents]) you could reliably determine if the child is black or Hispanic and therefore worthy of an adversity score boost. Everyone else, get stuffed. They may also be making a big show of using these factors that "technically" don't identify the test taker's race and instead are just disregarding them and relying solely on the self-reported racial information provided when taking the test. Who knows? But I highly doubt you'll find a lot of poor white kids in West Virginia getting boosted scores via this scheme.
> But I highly doubt you'll find a lot of poor white kids in West Virginia getting boosted scores via this scheme.
That is exactly what I expect to happen - as well as poor kids who happen to belong to racial minorities. The College Board published the criteria they will use to calculate the score. Race is not one of the criteria. But you're saying that race will not only be used, but be the primary criteria. So, let's be clear: you're saying the College Board is lying. Which, I can't prove they are not, but I'm inclined to trust them over your claims.
Just to nitpick a bit, race is one of the optional criteria. The school decides whether they want the College Board to include it.
"The program aimed to measure the challenges students faced. It created an expected SAT score based on socioeconomic factors including, if schools chose to add it, race." [1]
Thanks. CBS seems much better than the NYT on this.
Quote:
These factors are first divided into three categories: neighborhood environment, family environment and high school environment.
Each of the three categories has five sub-indicators that are indexed in calculating each student's adversity score. Neighborhood environment will take into account crime rate, poverty rate, housing values and vacancy rate. Family environment will assess what the median income is of where the student's family is from; whether the student is from a single parent household; the educational level of the parents; and whether English is a second language. High school environment will look at factors such as curriculum rigor, free-lunch rate and AP class opportunities. Together these factors will calculate an individual's adversity score on a scale of one to 100.
Thanks for the link. I would honestly be quite astounded if they were using individual-level reporting instead of using public data captured at the census tract level. Do you have any information about where the "family-level" data are coming from?
Realistically, this will be based on ZIP code and registered high school. Information that's already publicly-available. This changes nothing except potentially making neighborhood-level affirmative action more politically palatable.