Teachers get it handy!

[David McKeown]

Really specific scenarios like that are impossible to account for in modeling as there wouldn't be enough data to support it.  Again as I said in an earlier post, models do well at predictions across large populations but can be very bad at the individual level.

I accept I used unusual and somewhat niche examples to emphasise the point but I think the point still stands. Why in your scenario as I understand it (and again there's no confirmation of this from CCEA) is the class ranking of a school with a history of excellent results in a particular subject given the same weight as a school with a terrible history. As it would appear might Have happened here.

Another example from my school in any year we would have had about 20 maths students. The top 2 might have scrapped an A or more likely received a high B. A large part of the reason for that was I'll health of the teachers resulting in a less than ideal scenario of substitute cover for large parts of the year. I think in my year we had three. I know of plenty of schools who would have expected 12-15 A's from similar ability pupils including very strong A's. How is a system that awards me the same or more for my scrapped A fair when compared to what was awarded to the strong a in the other school that had a class ranking of 8th.

Some schools for whatever reason have a history of certain subjects outperforming other subjects. From what I understand that simply wasn't factored in even though in reality it would have had a significant impact on likely results. The consequence of that failure to factor that in is that some kids lives are dramatically altered.

[delgany]

From Education committee today , Minister Weir and CEO of CEA- Justin....., both stood over the need to moderate and standardise the results to protect the credibility of the exams.
England, Scotland & Wales have abandoned that plan, so does this not doubly disadvantage those whose grades were moderated downwards.
They continually referred to anomalies being addressed by appeals in batches by schools , across subjects and on that point, they anticipate a X 5 increase in appeals. For A2 students these have to be addressed before 7th of Sept !

[Milltown Row2]

From Education committee today , Minister Weir and CEO of CEA- Justin....., both stood over the need to moderate and standardise the results to protect the credibility of the exams.
England, Scotland & Wales have abandoned that plan, so does this not doubly disadvantage those whose grades were moderated downwards.
They continually referred to anomalies being addressed by appeals in batches by schools , across subjects and on that point, they anticipate a X 5 increase in appeals. For A2 students these have to be addressed before 7th of Sept !

Id imagine they they would need to be sorted out long before the 7th as college places would be going and less chance of being accepted, never mind finding digs/halls!

[johnnycool]

From Education committee today , Minister Weir and CEO of CEA- Justin....., both stood over the need to moderate and standardise the results to protect the credibility of the exams.
England, Scotland & Wales have abandoned that plan, so does this not doubly disadvantage those whose grades were moderated downwards.
They continually referred to anomalies being addressed by appeals in batches by schools , across subjects and on that point, they anticipate a X 5 increase in appeals. For A2 students these have to be addressed before 7th of Sept !

Have England backtracked as well?

[clawaddy]

I have taught GCSE and A level for many years. In my experience there were pupils who you could confidently predict their grade
- at both ends of the spectrum, but there were quite a few for whom it was difficult to predict. When the exams were cancelled in March I couldnt believe the promises that were made that grades could be awarded accurately and no pupil would be disadvantaged. It is a near impossible goal to achieve.

[CK_Redhand]

Really specific scenarios like that are impossible to account for in modeling as there wouldn't be enough data to support it.  Again as I said in an earlier post, models do well at predictions across large populations but can be very bad at the individual level.

I accept I used unusual and somewhat niche examples to emphasise the point but I think the point still stands. Why in your scenario as I understand it (and again there's no confirmation of this from CCEA) is the class ranking of a school with a history of excellent results in a particular subject given the same weight as a school with a terrible history. As it would appear might Have happened here.

Another example from my school in any year we would have had about 20 maths students. The top 2 might have scrapped an A or more likely received a high B. A large part of the reason for that was I'll health of the teachers resulting in a less than ideal scenario of substitute cover for large parts of the year. I think in my year we had three. I know of plenty of schools who would have expected 12-15 A's from similar ability pupils including very strong A's. How is a system that awards me the same or more for my scrapped A fair when compared to what was awarded to the strong a in the other school that had a class ranking of 8th.

Some schools for whatever reason have a history of certain subjects outperforming other subjects. From what I understand that simply wasn't factored in even though in reality it would have had a significant impact on likely results. The consequence of that failure to factor that in is that some kids lives are dramatically altered.

Any time somebody gets upgraded for a particular reason, somebody else has to be downgraded. That's the nature of standardisation as proportions of grades have to balance out to what CCEA deem acceptable.

Using or not using particular variables could be argued against as being unfair for those they disadvantage.  Exams are there to reward pupils, yes. They are also there to differentiate between higher and lower performing pupils. CCEA are in the position of maintaining "fairness" while also allowing a means of differentiating between them.  The two goals are really at odds with one another.

[FermGael]

Here's a specific case for you .
Consortium arrangement between Grammar and Secondary schools in Northern Ireland.
A Level Biology.
Grammar pupils got what the teacher predicted .
Secondary pupils , all predicted by the same teacher , got downgraded by on average one grade.
Just beggars belief

[CK_Redhand]

Here's a specific case for you .
Consortium arrangement between Grammar and Secondary schools in Northern Ireland.
A Level Biology.
Grammar pupils got what the teacher predicted .
Secondary pupils , all predicted by the same teacher , got downgraded by on average one grade.
Just beggars belief

It could still be explained by the model though. Their AS grade might have been lower. Also the standardisation by school adjustment is probably at play here.

Not saying this is right or fair, but with the raw data we would know how the model scores them.

[CK_Redhand]

On BBC newsline they have just said that school past pert was not used in the A level models but was used in the GCSE ones.  The confusion continues

[David McKeown]

On BBC newsline they have just said that school past pert was not used in the A level models but was used in the GCSE ones.  The confusion continues

Yes that was my understanding too. That's why I feel that certain variables which would have made significant real world impacts have been ignored and kids futures have been altered. That can't be justified by saying yeah you got screwed but bask in the warm fuzzy feeling that we have ensured consistency of results between last years exams and this years nonsense when you missed the last three months of school.

[CK_Redhand]

On BBC newsline they have just said that school past pert was not used in the A level models but was used in the GCSE ones.  The confusion continues

Yes that was my understanding too. That's why I feel that certain variables which would have made significant real world impacts have been ignored and kids futures have been altered. That can't be justified by saying yeah you got screwed but bask in the warm fuzzy feeling that we have ensured consistency of results between last years exams and this years nonsense when you missed the last three months of school.

It contradicts what CCEA have on their own website though from johnnycools link.  Also some school principals have said they were sure it was used.  Some clarity would be nice

[ONeill]

A Level grades are based in the main on AS grades with some uplift from expected AS repeats.

AS grades are based in the main on average GCSE grades.

GCSE grades are based in the main on 3 years' outcomes from same school, with each year weighted (2019 results given heavier weighting).

Statistical tests have shown a high correlation when 2019 results were tested. Though high could be close to moderate and isn't very high.

This was explained on here before the results came out.

No real confusion at all.

[ONeill]

Every model had its problems.

My solution was also riddled with problems but I'd have done this:

Heads of Depts would have presented their predictions to principals and senior management. They would take it away, look at it, look at all the variables within the school for that cohort and previous years and see if they could professionally, and with integrity, stand over them. Feedback to the HoDs individually and adjust if necessary. Boards spot-check random schools to confirm consistency.

This was close to what happened but it was watered down and the power was left in the hands of non-teaching civil servants to decide.

It's a bit like football club owners dictating who managers should buy.

[Milltown Row2]

Here's one for you to debate, didn't know whether to put it in the WTF thread..

A Belfast integrated school teacher (life long friend) said she got all A.* for her students even though she'd predicted a handful of B's and rest A's... No A*s were predicted

Now it's BTech level but still equivalent to A level grades!

[ONeill]

Can you give Bs and As in BTech?