Grading fiasco turnaround - it's about time!

[Schmedz]

Although it would obviously have been better if the government and Ofqual had got things right in the first place, the news about the U-turn on A-Level and GCSE grades is such a relief!!

Fair treatment for deserving students everywhere.

I feel now for all the schools who will now probably have to reprint a whole lot of award notices for their students on Thursday (and having to rearrange information at the last minute, yet again)... although DD's school sent an email earlier today withdrawing the arrangements for the students to come in and said an email will be sent at 8am instead. School will be open for any wanting to appeal or who need to discuss A-Level options based on their grades ...

Pretty sure DD will be wanting to see some close friends to celebrate now! And oldest DD is delighted that she now gets her A* in one subject which was downgraded (thankfully she got her first choice uni anyway - we've been confirming accommodation today and feeling very excited).

Cross fingers for all with GCSE DC this week

I think the mistake was allowing teachers to actually specify a grade at all in the predictions submitted. they should have just got the teachers to rank the pupils in order of expected performance within each subject just numerically, numbering from 1 to (number of pupils entered). This fiasco is so unacceptable because of huge numbers of situations where e.g. teachers predicted a B and a C was awarded (same goes for As to Bs etc) - this was easily foreseeable. Within every cohort where there are a number of pupils taking an exam who are capable of achieving a B grade, some are going to randomly cock up, fail to read the question, get a bad night of sleep or for some other reason not show their full ability on the day, and only get a C. Obviously the teacher can only predict a B for all of them, obviously they don't genuinely expect the whole group would have got a B in a normal year but there is no way to assign which ones would have cocked up. with no correction there is massive grade inflation leading to devaluation of all qualifications. This "normalisation" process attempted to simulate this effect and it's not fair or right but the alternative they were trying to avoid is that which we will see instead - in future years employment recruiters will know that an unusual number of A and B grades got awarded in 2020 and will treat all such results with suspicion. this year's cohort will need to prove themselves all the more rigorously and even the ones whose A and B grades were never in doubt will suffer for it as when the chips are down, a recruiter will doubt whether that grade is truly deserved.

if instead just a numerical rank was submitted, they could have then used past results to translate that data into rankings across all schools, and only then do they create grade boundaries setting them in such a way as to ensure only minimal grade inflation. it would still be a deeply flawed system but at least there would be no specific 'downgrades' causing such heartache.

@YinuCeatleAyru
Do employers really set such store by A level grades?
My dh is an employer and he certainly doesn't set any store by them. A levels are used to get into uni. The uni grades in 3 years' time will be as they ever were. If, hypothetically, someone got 3 A at A level but gets a poor 2:2 or 3rd class degree then the degree will override those A levels. DH says he barely glances at A levels and when employing people is far more interested in their skills and experience. He says he would not automatically employ someone with, say, AAA over someone with BBC - those results would make zero difference compared to aptitude, skills, experience and attitude.

The problem is that your method forces teachers to absolutely rank young people when in reality that isn't possible. Yes top and bottom are easy but ranking who is 5,6,7,8 becomes pretty abitrary, and it's not a guarantee that on exams day person 6 will definitely do better than person 8 because there is, in reality, a hair's breadth between them.

I guess it will be variable. certainly if recruiting graduates with work experience one grade will make no difference. The recruitment I did for 10 years (no more thank god) involved selecting 20 recruits from typically 300 applications, all of whom would be kids with no degree or experience but fresh out of A-levels and now in their first or second year of university seeking a part time work opportunity which was well paid and would open doors to an excellent career down the line. and they were all bright and enthusiastic and hard working and there was very little to distinguish between them, one had to narrow the field down somehow and certainly when distinguishing between two otherwise identical-seeming candidates, an A-level grade could cinch it. not for the final job, but for selecting which applicants to put into an interview round. I do think that for recruiters in that kind of situation where they are needing to find something to nitpick because there isn't time to interview all the excellent candidates and it comes down to looking at grades, a B from 2020 is going to be worth less than a B from 2019.

you are right that my alternative suggestion was flawed, as I acknowledged above. The "real" results that are awarded in a normal year have an element of chance to them which is impossible to simulate with any kind of algorithm in a fair way.

@BwanaMakubwa

Exactly right!

And many centres already considered past results in their ranking knowing that if they overestimated unrealistically the algorithm was likely to drop some below their CAG - this is why the ranking was also used (the videos sent out before results explain it clearly). The problem is trying to make a ‘normal distribution’ from an unusual year.

Moderation could have been done by random samples of work from each centre to check how accurately the CAGs were awarded.

Better 12% grade ‘inflation’ than 100% of a system in disarray where privileged students who were in small classes were awarded their CAGs and those in larger classes/historically poorly performing schools were severely downgraded.

Best scenario out if a number of imperfect scenarios (let’s not get started on how linear exams themselves aren’t particularly brilliant wats of assessment either...)

Schools Still Used the Abandoned Algorithm.

Up to 1,900 schools still used a similar algorithm to the one the government abandoned due to "too many significant inconsistent and unfair outcomes". 1,900 schools sent their CAGs to FFT Datalab who sent back to them, information about how many grades to change to fit their school's 2019 data. 1,000 schools sent them to FFT again to further check their CAGs.

I'm writing to FFT, my MP, Ofqual, my school, exam boards, any journalist whose written about the algorithm. I suggest you share this information and do the same. My child's grades were at or lower than his most recent sustained, attainment grades.

Grades should have been, per Ofqual, "[a CAG] which represented the grade that student would have been most likely to achieve if teaching and learning had continued and student had taken their exams as planned."
AND
"Working At Grades" as well as "Target Grades" (FFT Target Data) were not to be used as per Ofqual, "note that these centre assessment grades [CAGs] are not the same as - working at grades (the grade a student is currently working at) [and] - target grades". Rather, the DfE and Ofqual stated when determining CAGs to use "student's knowledge, skills and abilities in relation to the subject. This evidence should inform teachers' professional judgements about each student's likely performance at the time of the exam."; "those judgements should be holistic, based on the range of evidence that schools and colleges have. So students shouldn't worry about one disappointing mock exam result"; "performance over the course of study"; "review data, classwork, bookwork"; "performance on any class or homework assessments"; "signed off by at least 2 teachers [or an equal] in that subject .... [stating] fairly represent the grades".

Please see the following: ffteducationdatalab.org.uk/2020/06/gcse-results-2020-a-look-at-the-grades-proposed-by-schools/?fbclid=IwAR1i6OB4gEvjEcBa4yCWW9xYW83S5EM1K2WiWanh-BKoqTpo1Q0V2Z9-V1g

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/909035/6656-2_-_Executive_summary.pdf

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/887018/Summer_2020_Awarding_GCSEs_A_levels_-_Info_for_Heads_of_Centre_22MAY2020.pdf