Some grade boundary stuff:
Cblue ^Then they take whoever is left in the group (ie those that would have a 7 or 8) and give half of them a 7 and half of them an 8.
They then do their same exercise for grade 4-6, and then 1-3.^
This isn’t quite right. They peg the grade 1, 4 and 7 grade boundaries so that the proportions getting above those grades match those at G, C and A last year, and set the 9 grade boundaries using the formula. The grade boundaries for the other grades are then set at equal distances between these grades. So if it works out at 70% for a 7 and 80% for a 9, it’ll be 75% for an 8, regardless of the distribution of number of students between 70 and 80%. If it works out at 49% for a 4, and 70% for a 7, then it’ll automatically be 56% for a 5 and 63% for a 6. If decimals then there’ll be an argument about which way to round.
The formula for a 9 is “Percentage of those achieving at least a grade 7 who will be awarded a grade 9 = 7% + 0.5 × (percentage of students awarded grade 7 and above)”
Whether grade boundaries are higher or lower depending on how bright the year group is - technically the grade boundaries should work out the same for any year group (although unlikely, but they do a lot of statistical analysis of subgroups to make sure they perform as expected). If the year group sits an exam and the 4 grade boundary is set at 40% because 50% of the students should pass, then if a brighter cohort sat the same exam, if the grade boundary was set at 40%, more than 50% of students would pass, which is what you’d expect.
I don’t know how grade boundaries are set for IGCSE, they can’t use KS2 data because a lot of students sitting IGCSE don’t sit SATs. Maybe they actually read the students’ work and set the standards based on that? 