Althouse: "Even after a final term with schools closed for the pandemic, Sam Sharpe-Roe was optimistic about the coming school year."

August 20, 2020

"Even after a final term with schools closed for the pandemic, Sam Sharpe-Roe was optimistic about the coming school year."

"Teachers from his West London school had given him grades — three A’s and one B — that were strong enough to secure him a spot at his first choice of university next month. But after the British government used a computer-generated score to replace exams that were canceled because of the coronavirus, all his grades fell and the college revoked his admission. Mr. Sharpe-Roe, along with thousands of other students and parents, had received a crude lesson in what can go wrong when a government relies on an algorithm to make important decisions affecting the public.... Nearly 40 percent of students in England saw their grades reduced after the government re-evaluated the exams, known as A-levels, with the software model. It included in its calculations a school’s past performance on the tests and a student’s earlier results on 'mock' exams."

From "British Grading Debacle Shows Pitfalls of Automating Government/The uproar over an algorithm that lowered the grades of 40 percent of students is a sign of battles to come regarding the use of technology in public services" (NYT).

24 comments:

Michael K said...: Affirmative action not affected.; 8/20/20, 4:06 PM
rehajm said...: It isn't so much the algorithm as the doofuses in government not understanding their limitations and forming policy that strictly relies on their outcome.; 8/20/20, 4:08 PM
NotWhoIUsedtoBe said...: The ecological fallacy in action.; 8/20/20, 4:15 PM
Expat(ish) said...: Will they reach the obvious conclusion? That a sprawling LCD qualify government shouldn't be in sole charge of hospitals, schools, and universities.

-XC; 8/20/20, 4:15 PM
cubanbob said...: There never was a legitimate reason to cancel the tests. Of course these nimrods didn't take their conclusion to the next logical step; if they need an algorithm to correct the test score then how legitimate are online tests?; 8/20/20, 4:20 PM
mikee said...: Which one used race as a part of the decision, the human scoring method, the algorithm run by computer, or both? And if both, did they both calculate racial inputs the same, or differently, for any given skin tone?; 8/20/20, 4:20 PM
NotWhoIUsedtoBe said...: The other problem is the students whose scores were increased. They didn't win fairly.; 8/20/20, 4:21 PM
Joe Smith said...: Read the article and didn't see the answer (but I went fast)...

If he's white then he's fucked.

England is dead...might as well be Islamabad.; 8/20/20, 4:24 PM
Churchy LaFemme: said...: That a sprawling LCD qualify government shouldn't be in sole charge of hospitals, schools, and universities.

*LCD* quality?

Nixie Tubes at best.; 8/20/20, 4:30 PM
MadisonMan said...: Free version of the article (for you French-speakers. Or the google-translated version here.; 8/20/20, 4:33 PM
Fernandinande said...: an algorithm that lowered the grades of 40 percent of students

AKA, "Grades of 40 percent of students had been inflated."

But students and their parents, particularly those from lower-income areas with struggling schools,

AKA, "particularly those from lower-income areas with stupid students and documented grade inflation."

"These algorithms are obviously not correct,” said Mr. Sharpe-Roe,

I guess he's the expert at education testing and statistics, what with being 18 years old and BLACK.; 8/20/20, 4:35 PM
Josephbleau said...: What stuff like this does, Ivies included, puts more talented people into the lower ranked schools. When the lower ranked school kids do better in grad school and in job performance, the high class schools will be nocked down a peg. Many large companies recruit only from selected state flagships now for undergrad. MBAs are next.; 8/20/20, 5:05 PM
Jupiter said...: Unless the "algorithm" changed the number of slots available, it's a wash. Some got in, others didn't. Same as if no grades had been increased, reduced, inverted, or otherwise tampered with. The fact that the idiots at the NYT, who presumably attended a university for four years, can't grasp this simple fact indicates how little difference it makes. You can't educate a sponge.; 8/20/20, 5:10 PM
mandrewa said...: As numerous people have commented or are going to comment, there was no rationale reason for the schools not to do the tests anyway, despite Covid-19 -- well except for opposition from the teachers union and the public school administrators.

So instead they assigned these life-altering scores to students based on how they had scored on practice exams in February or January plus how other students in the same school had scored in previous years. Or in other words, if you are student under this system, your score is determined not only by how well you do on a test but it also depends on how well the other people in your school did. If they do poorly or had done poorly that will lower your score.

It sounds terrible, but guess what?

There's an article in the Electronic Telegraph today where someone that should know revealed that actually they've been doing this for the last ten years. That is people are assigned grades not only based on their own scores but on how well other people in the same school did.

It really sounds wrong. Naturally I wonder why they would be using such a seemingly irrational system to allocate scarce resources, ie. placement at top universities.

It's probably unsayable in the United Kingdom, maybe even literally a crime, but I do wonder if there are additional factors, hidden in this inscrutably complicated algorithm, that might also be being used to assign grades.; 8/20/20, 5:15 PM
wild chicken said...: Sounds like the algorithm didn't give any break to minorities. Oops!

Oh well, they tried.; 8/20/20, 5:45 PM
readering said...: The editorial in the conservative Telegraph said it all, “the exams fiasco in England beggars belief, given the time authorities have had to prepare.”

In Britain, universities offer conditional places based on the expected grades on national A level exams taken at the end of final year in 2-4 subjects studied over a 2 year period. These exams are not computer-graded multiple choice, but arduous multi-day affairs with outside graders. When I took A levels in the seventies in 3 subjects (with S level in 2), it came to a total of 10 exams over a 3 week period in July. At some point after my day some A levels included a course work component, although I believe that has been cut back recently. All the results are published on the same day in August, and if students don't match the grades required by their first choice uni they either have to plead for their lower result to be accepted or scramble to find a different uni that will accept the grade based on vacancies created from students taking a better offer. It's a messy process even in a normal year. Keep in mind that students are applying for entry into a specific department at the uni, although one can later try to change departments. All the other factors that go into college applications in the US do not apply. Oxford and Cambridge also conduct interviews to whittle down applicants, but I don't believe other universities generally do that. One can also apply from other countries, including the US, using their systems, like the Bac in France.; 8/20/20, 5:54 PM
rhhardin said...: It's no different than having your name tossed in a hat. Like the retirement plan of a failing company. It's fair, just stupid.; 8/20/20, 6:58 PM
Michael K said...: As numerous people have commented or are going to comment, there was no rationale reason for the schools not to do the tests anyway, despite Covid-19 -- well except for opposition from the teachers union and the public school administrators.

My grand daughter is driving to Tucson next Friday to take her SAT. All the SAT sessions in CA have been cancelled and she is a senior.; 8/20/20, 6:59 PM
Joe Smith said...: @Michael K

Safe travels...; 8/20/20, 7:36 PM
Yancey Ward said...: Some people didn't like being assigned to Slytherin.; 8/20/20, 8:18 PM
Bunkypotatohead said...: Why bother to attend the school at all?
Just let the Gov't estimate what everyone's grades would have been if they had gone there. Would save a lot of time and money.; 8/20/20, 8:52 PM
Vonnegan said...: Michael K - we're going to Waco for the SAT at the end of the month because the CC there tells me they're definitely having the test no matter what. No one in Houston wants to host it, including my son's private school. Cowards. We'll probably wind up doing the same for the 9/12 ACT, and who knows about the 9/19 SAT? My son has no scores, and coming from a good private school, he's going to need them. Admissions are going to be ugly this year; I can't wait to watch UT reject every applicant from every private school in Texas who isn't in the top 6% of their class. Diversity!; 8/20/20, 10:10 PM
A10pilot said...: My daughter is currently finishing up her Masters thesis work at Durham University so I asked her about this. Her response:

“They’ve completely reversed it. Everyone got their projected grades. So now the uni’s are screwed because they don’t have enough spaces. They’re now begging people to defer by offering money off their tuition.”; 8/21/20, 8:25 AM
Thomas W said...: If the computer algorithm will potentially adjust everybody's scores, you'd expect about half to have their grades lowered and half raised. Is this 40% some sort of bias or just random chance?; 8/21/20, 1:00 PM