Exam results

Well only those not going onto further education actually need them. And those going into the work force don't have some of the issues that you have raised with them resitting them.

And you keep mentioning fair but what is fair about them having inflated grades. How is that fair to the people from last year who may be competing for the same jobs, how is it fair to students whose teachers graded them fairly?
Look at @Keefly Bantam post about his son. He says the teachers grades were lower than what the algorithm produced. That means he is in that 16% where teachers predicted fairly (or maybe even in that 10% where teachers under predicted). He is actually being penalised for his teachers being fair because someone less intelligent than him from another school whose teachers over predicted will now have the same or better grades than him when they didn't deserve it. When you are dealing with opinion and every single persons opinion will differ you need a way to normalise that which is what the algorithm did, it adjusted the schools results on a curve based on past results, basically if a school over the last 5 years has an average of 6 A* pupils and this year it had 12 based on predictions its adjusted down as its highly unlikely they had this 1 off group of kids who are more intelligent than previous years. There will be people this screws over especially as the inaccurate predictions play a part but it screws over less people than having these inflated predictions and the lack of conisteny on how these predicted grades were obtained.


And again on the prep time. They already had this prep time. Infact A-Level students have also already had it at GCSE as well and unless its a subject they didn't do at GCSE and a subject that has a different way of presenting questions there isn't going to be a massive difference in styles. The only thing they are really missing out on is being walked through what is likely to be the actual contents of the questions IF THE PAPER CHANGES. But they wouldn't have got that if they had been doing resits either and people get through resits perfectly fine.

 

To second @Aaron Baker what exactly are the other options that would normalise predicted grades that are done on opinion rather than hard numbers. I'm genuinly intrested as I can't see how you can adjust for some schools over predicting more than others without it being normalised on a school level rather tha national one that would normally happen.

Why is it fair that pupils at a school where the teachers who try and be fairer to a point they slightly underpredict how many A's and A* they will get in a normal year now find its pupils doing worse than a school who usually get fewer but because they over predicted are now getting more.

 

But they were downgraded because their school/colleges last 5 years results suggest they wouldn't get as many A's as they predicted. As the results were up even with the algorithm it suggests there was some margin of error to allow them to have a few more than they statistically would but in many cases they are up signficantly. As @Aaron Baker stats he posted show A and A*'s are up 6.1% and overall pass rates are up over 10%. That's a sigfnicant jump.

Again the problem isn't with the alogirthm, its with teachers for whatever reason being prone to over predicting the number of A and A* and then if a nailed on A student gets downgraded potentially getting ranking wrong (perhaps due to a non-intelect based bias)

 

I came to the conclusion via two things

1) The studies don previously on the accuracy of predicted grades which is something Uni's have been bitching about for years. The 16% comes from one such study

2) The increase in both those getting top grades and the pass rate. They are vastly highly that the usual rise. I believe its usually an increase of both by around 1.5% and we are seeing over 6% for A/A* and over 10% on the pass rate

If predicted grades were done on just hard numbers you wouldn't see them being at grades higher than they got in their exams/tests. I've seen countless people on the news and online when discussing this saying how they were got one set of grades in their mocks, were predicted something higher but then got something closer to their mocks in the algorithm. We also wouldn't see them be so far out, yes there will be some people who just flop under the pressure of the actual exam and some who excel under the pressure of an actual exam but only 16% being right?