To think that this is discrimination

[unnervingrabble]

The BBC have reported that 'The Scottish Qualifications Authority lowered grades using an algorithm - with pass rates for pupils in deprived areas downgraded further than those in more affluent parts.'

Surely this this discrimination against poorer Students who have been doing Highers? They have had their grades reduced by 15% in the most deprived areas but 7% in the most affluent areas.

Finally got back to the thread!

But the problem is that young people who are from less aflluent areas do worse. Always have done worse, for a whole host of reasons.

Some young people from less affluent areas do worse. Those who are from less affluent areas who were predicted to do well based on the P7 tests and results in the Nat5 exams will have lost out because of this algorithm.

"Artificial Intelligence" identifies systemic biases as meaningful patterns, and amplifies them.
Garbage in, garbage out.

It's a well-known problem in computing circles, that somehow politicians and other decision makers always seem to be blind to, when rolling out automated decision-making systems onto the general population.
Same problem with computer vision algorithms, or AI predicting risk of reoffending (very "Minority Report"... only in the states, so far, thankfully), or opaque algorithms computing insurance premiums, or credit scores...
Discrimination is baked into such systems. It's well documented, and shouldn't come as a surprise.

as exactly where the problem lies. It is not in the statistical process applied to datasets but the methodology employed in building them that is cause for concern. Well, there has to be some method, in this case a bundle of methods, an exam board can use when exams cannot be taken.

What else would you suggest happens? A combination of all previous work can be/is used, any combination of coursework, mocks, tests, past achievements, estimated grades and an equation (that being all the algorithm is) to work out the best guesstimate. What else do you suggest.

My issue isn't that they write an algorithm - there has one been in use for a couple of decades now - but that the weighting of it seems to be at odds with UCAS data of many years.

And yes, the English boards will do something similar. Wales something a little different because of their differing exam system!

@CuriousaboutSamphire

I wasn't very clear - collecting the data (as you suggest) is fairly straightforward - assessments, exams, coursework grades, etc but I meant statistical analysis of that data is not the problem - but the methods employed at the next stage.

www.bbc.com/news/uk-scotland-scotland-politics-53719477

Also on Twitter twitter.com/stianwestlake/status/1292892098149199877

The problem is that if a school had, say ten pupils predicted to get five A’s based on prelims and coursework , it’s more than likely than maybe 5 or 6 will get those A’s in an exam setting. The other four may not have studied, had a bad day on the exams or a million other reasons why they didn’t do as well as hoped. And they would be bitterly disappointed, but ultimately would only have themselves to blame.
This year, It’s not their fault, they couldn’t have done better in course work or prelims and despite this have not got the grades the wanted.

What’s the answer though? Give everyone the marks they wanted? Which would mean something like a 20% + increase compared to every other year.

The most workable solution I can see is that the universities or colleges can let them into the course they had hoped to get into. Given the number of foreign students next year is likely to be way down, hopefully there would be spare capacity.

Given the number of foreign students next year is likely to be way down, hopefully there would be spare capacity.

That's one aspect, but there could be higher demand from students who would otherwise have gone on gap year travelling or working but now going to uni because gap year options have all but disappeared. Also students who were intending to do apprenticeship schemes may now go to uni instead as so many employers have suspended recruitment. There are a lot of factors causing swings each way and the net effect is hard to predict.

@serenada ah! Cross purposes.

"Artificial Intelligence" identifies systemic biases as meaningful patterns, and amplifies them

Absolutely. So for example, a pupil from a poorer area might need longer to work towards peak performance on exam day and not be at their best in their prelims or shine in previous assessments.

I know this happened to me throughout my education in Scotland.

Scottish education, in my experience, has been poor for a long time. 15 years ago, my school in a big Scottish new town was unable to offer me O Grade History, but as it was only an O Grade, I was permitted to study it myself in my own study hours, and I got an A. No teacher input whatsoever - I was provided with the materials and studied on my own in the library. History was again unavailable at Higher, and the 2 neighbouring schools who were meant to offer mainstream subjects that werent available at their partners, wouldnt offer it either due to low interest. This time, I wasnt allowed to study it in my own time, no matter how much I pled, and that was one much needed Higher for university entry that I didn't take. I had to take Art Higher instead, which I was fairly useless at and got a C for.

My mother went up to see the Headmaster of my school to complain, and was told he "didn't expect any pupils from this school to go on to university". It was shocking (and several of us did).

I went to see the careers advisor in school at age 16 or so and you had to pick from a range of future careers in different bandings. Having mostly As and Bs and being quite ambitious, I selected from the highest banding which contained careers such as lawyer, teacher, lecturer, doctor, dentist, etc.. The careers advisor for some reason got really annoyed at me for this and tried to force me to select a lower banding, getting very specific about the merits of librarianship in particular and telling me where I could go to study for that. It turned into quite an argument, and she was very intimidating towards a 16 year old as I would not back down on my choice. Looking back, it always baffles me that someone in education should try quite so hard to stifle the ambitions of a clever 16 year old. She basically told me that I was aiming too high, I wasn't clever enough and I would end up disappointed. I went on to excel at university but due to the lack of Highers, I did have to study English for a year first and then transfer to Law based on my exam results at university. I think by the time I had got to university, I was so used to independent study and motivating myself that the transition from school was unusually easy.

The exam hall roof developed a serious leak on the day of my Modern Studies higher and we could only sit crammed into one side of the room, listening to dripping sounds. I still got an A.

But in my prelims I was a full grade lower for everything. As a 17 year old, it took me those extra few months to develop the steely focus and determination needed to perform better in the exam itself by studying more.

I developed a realisation as a child that the Scottish education system was not necessarily there to help me fulfil my ambitions and make the most of myself. Really, really odd and I suspect it still happens quite a lot.

The most workable solution I can see is that the universities or colleges can let them into the course they had hoped to get into

The problem with this is that most children who have been disappointed with their Higher grades haven't applied to Uni yet. In ordinary times this is a key advantage of the Scottish system - most Uni offers are made on the basis of exams taken in the penultimate year at school. So the people who were predicted AABBB by their school and are now sitting with BBCCD haven't even started that application process.

Different from the English/Welsh/N Irish situation where kids sit A-levels in the spring, already knowing what they need for a Uni place the following September.

Who knows what happens? If they do (another) U-turn and just go for teacher grades for everyone who was downgraded, there will be a lot more students with the grades needed for Uni places. Will Unis just up the requirements? So for a 2018 entry into a course you might have needed AABBC, next year they'll be looking for AAAAB from the people who "sat" highers in 2020.

@unnervingrabble

It'll be interesting to see if the same discrimination occurs with the A level results in England.

In cities that would need to be done on a school by school basis - you can get private close to failing state schools.

In Scotland Swinney and Sturgeon also have the education strategy 'closing the attainment gap' which they been promoting for years, but at the same time seem to have conveniently forgotten about when they described it last week as 'not credible' that poorer children would have got good results.
They don't give a stuff about children really. Or the teachers and local authority education boards who have triple checked the credibility of their predicted grades before submitting to SQA.

Of course they don’t - all mouth that lot.

I've got 2 perspectives on this - they will seem contradictory. Ugh.

A family member is having to appeal 3 results. They have performed exceptionally all through school - winning prizes etc.

Conversely I tutored someone for English and asked another family member for help (senior English teacher, private school, trained exam marker too). I shared the work by this person and was told by family member - it's a pass but not an A.

The school gave an A!

So schools do mark optimistically

I would love to know how many students (ever) have actually achieved the same results that they were predicted. My own DD would have done far better overall if she was awarded her predicted grades. In my opinions exams are not the best way of working out ability. I know loads of people will disagree, but I believe they only test what you can REMEMBER, under pressure, NOT what you have learned. I am old and did O Levels, so no course work involved. I left school with precisely 1 -at grade C. I then did a level 3 Btech type course and now have a level 5 teaching Diploma that was assessed through course work and teaching ability. I know this is not the case for everyone, but for each person who will feel penalised through 'predicted ' grades I am sure there will be one who will fare better than if they had to sit exams!

I would love to know how many students (ever) have actually achieved the same results that they were predicted.

Well I got more or less what I was predicted/expecting when I did O and A levels back in the 80s.

My son got virtually exactly what he was predicted in his GCSEs two years ago. We'll know in two days time whether he'll get his predicted A level grades.

I'm too old - i can't remember ever being told predicted grades.

Also agree with the general points made by @IncandescentSilver. I was at a very ordinary state comprehensive in Scotland through the 1980s. Nobody from my school, even the most academically able, were pushed hard to apply for Oxbridge. That just wasn't for the likes of us. Out of my year of about 250, one went to do vet medicine, a couple went to study law. Again, those sorts of subjects weren't for the likes of us. Leave that to the private school kids, or perhaps Gillespie's kids, at a push.

@user1497207191. With the greatest respect 'more or less' and 'virtually' are NOT 'exactly the same grade as predicted '!