https://www.chronicle.com/article/A-Tech-Whiz-Is-Conquering/244305?cid=at&utm_source=at&utm_medium=en&elqTrackId=1d5ec81168f04ebe9e9c8ea8e4e30c8c&elq=37d52599534943e08dbdd213e1acd4af&elqaid=20234&elqat=1&elqCampaignId=9470
(not sure if the link will work. It is ungated, title is “A Tech Whiz Is Conquering College Admissions. It Takes Charm, Innovation, and Dancing Sharks” published this week.
This article confirmed for me the data-crazy environment of today’s admissions. I am most interested in learning more about the predictive models that the schools use, whether prospective students are aware the data that the schools have on them, and whether the schools are concerned with the “error rate” of their models. This article confirms that some schools track whether or not an email is opened up and read, how long afterwards does the student go to the web-page, etc. Creepy, not cool. What if a student uses high levels of computer security that would prevent that sort of information from being tracked? smart kid, but the school might ignore because no data… and what about when the prospective students start using bots to give the appearance of interest (at least electronically). As a parent this alarms and infuriates me.
Salient quotes from the article (the program is named Slate)
“People are still buzzing about the “configurable joins” feature, which will allow users to combine just about any data point with another, meaning an admissions officer can easily pull up a list of, say, college counselors, the names of incoming students associated with each counselor’s high school, the median grade-point average for those students, the most-popular extracurricular activities of those students, and the total number of those students’ siblings still in high school. Cool, right?”
…
“Slate allows an admissions office to see which emails students open, which web pages they visit, and how much time they spend there. (“I’m not a stalker,” one enrollment official here says, “but my CRM is!”) Ping, the platform’s analytics service, matches up everything a user does on a college’s website to other information in that user’s Slate record.”
…
“Call up any student’s file in Slate, and you’ll see a series of colored dots, each representing some form of engagement with the college. The student attended this college fair on this day, clicked on this email link at 1:13 a.m., five days later — it’s all right there. Many colleges fold measures of a student’s “demonstrated interest” into the statistical models they use to predict who will enroll. Those data points can inform decisions about who gets a viewbook or, on some campuses, who gets an admission offer.”