Slashdot Mirror
University of Arizona Tracks Student ID Card Swipes To Detect Who Might Drop Out (theverge.com)
An anonymous reader quotes a report from The Verge: The University of Arizona is tracking freshman students' ID card swipes to anticipate which students are more likely to drop out. University researchers hope to use the data to lower dropout rates. (Dropping out refers to those who have left higher-education entirely and those who transfer to other colleges.) The card data tells researchers how frequently a student has entered a residence hall, library, and the student recreation center, which includes a salon, convenience store, mail room, and movie theater. The cards are also used for buying vending machine snacks and more, putting the total number of locations near 700. There's a sensor embedded in the CatCard student IDs, which are given to every student attending the university. Researchers have gathered freshman data over a three-year time frame so far, and they found that their predictions for who is more likely to drop out are 73 percent accurate. They also have plans to give academic advisers an online dashboard to look at student data in real time. "By getting their digital traces, you can explore their patterns of movement, behavior and interactions, and that tells you a great deal about them," Sudha Ram, a professor of management information systems who directs the initiative, said in a press release.
The article is void on information on what specific statistics indicate a student is more likely to drop out. Are students who use their ID card to go to the rec center more likely to drop out over students who us it to enter the library? The article doesn't say.
Okay, I was going to dump on this, because the TheVerge article sucks. The press release, however, actually does a good job discussing some of the signals they track and how this ties into them. They even have a nice visualization of student traffic which hints at some ways that they might be able to infer stuff from all of it.
As an aside, the article contains this horrible quote (I really hope there's some missing context):
We think ...[we're] sort of doing what Amazon does — delivering items you didn't order but will be ordering in the future
I'm sorry, but I do not recall Amazon ever doing that. Quite frankly, I'd consider it really awkward to receive things in the mail based on what they thought I might need.
Take it to the limit, everybody to the limit, come on, everybody fhqwhgads.
Bayesian solutions should be capable of >80%.
The alledged wisdom of the crowds should get close to Bayes.
73% is a miss. They should take a class.
Read the article, it talks about that too. The article claims the $/student is higher than in the 60s, but dropped somewhat from the 90s.
Also the article claims that the administrative layers became bloated in the universities.
This match what I see in practice; though I haven't crunched numbers.
There is less money per student than there used to be; I wasn't around in the 60s but it certainly feels like there is funds than 20 years ago.
Also a lot of funds these days go to what I would call non academic expenses like a gyms, student health centers, on-campus dining options. While I understand the value of these, they pull money away from running classes.
The number of administrators we have today seems also a lot higher than it used to. I am not always sure what the administration actual contribution is; it is hard to tell.
I'm surprised that the researchers didn't start to run into university politics before they published their result -- namely, that the University is using data to segregate students and preferentially help some students and not others.
Many a data science / predictive algorithm study has been sunk because university administrators think it singles out people, even if it is to help them.
I didn't say it was perfect, but I am pretty sure that someone who gets a 25th percentile SAT score ain't gonna last too long at a top-25 university, no matter what rocks-for-jocks major he tries to slink through with. It won't catch the 1600 SAT nerd who drops out after two years to start a billion-dollar business, and it won't catch the type-A overachiever who has a mental breakdown after three semesters of 72 credit-hours and zero sleep hours a pop, but standardized tests have their place, along with grades, and other stuff that makes its way onto a college application in gauging who's more likely to succeed and who isn't.