Bromster
Badges: 0
Rep:
?
#1
Report Thread starter 6 years ago
#1
I don't really know what's happening, they go from allowing 58% to score a grade 2 to 35% without any apparent reason. Like the SII grade boundary last year for a 2 was 67 marks, with 35% obtaining it. In the same year, the SIII grade boundary for a 2 was 48, with 58% obtaining it.

I managed to get in 3 full solutions and one partial, but I don't even know if I'll make my grade 2 offer. This sucks
0
reply
ctcmichael
Badges: 0
Rep:
?
#2
Report 6 years ago
#2
If you have 3 full solutions you are going to get a lot of marks, so I think you can almost guarantee a grade 2
0
reply
ctrls
Badges: 9
Rep:
?
#3
Report 6 years ago
#3
It's because setting the difficultly of the paper to be 'exactly right' is very difficult, so there's a huge variance from year to year. Some years they may end up making a paper too easy, or they may find too many people are getting into a certain boundary.

Also, the mark boundaries for III are generally lower than II, which are generally lower than I. That's just because of the relative difficulty of the papers, so they have their ways of balancing it.
0
reply
ETRC
Badges: 17
Rep:
?
#4
Report 6 years ago
#4
III is harder than II which is harder than I.
Harder papers have lower boundaries.
And 3 solutions is already 50% of marks so you will get Grade 2 normally.
Usually it is 5 for S, 4 for grade 1 and 3 for grade 2 in STEP II and STEP I.
STEP III is like 4 and part for S.
0
reply
hi1345
Badges: 3
Rep:
?
#5
Report 6 years ago
#5
I don't think people who replied to this have even read the OP, hes saying that Cambridge has changed the cumulative percentage of people acquiring a certain grade each year, while it should in fact stay about the same. So its very possible (actually it happens a lot) that one year there's an easy paper which has low boundaries (e.g., when 60% get a 2) and one year there's a hard paper with high boundaries (e.g, when only 40% get a 2)
1
reply
metaltron
Badges: 11
Rep:
?
#6
Report 6 years ago
#6
I think the wild variation in the percentages (of grade 2s) might be because grade boundaries are set using only the data from Cambridge offer holders (for II and III). Maybe somebody can verify this, but I'm fairly sure this is what I heard at the STEP school.
0
reply
Stray
Badges: 6
Rep:
?
#7
Report 6 years ago
#7
I think it's largely to do with the increase in numbers of students sitting the paper. Because it's an entrance exam, not a qualification, what matters is not the percentage of students who achieve each classification, but the number of students who are places in each classification.

I'm just crunching the numbers so that you can see it in data... 5 mins...
0
reply
Stray
Badges: 6
Rep:
?
#8
Report 6 years ago
#8
Here are the figures for 2007-2013, broken down by paper. I think it paints the picture more clearly than the %ages at each grade boundary from the examiners' reports.

STEP I
YearEntries1 (%)2 (%)1 (no)2 (no)
200756516.545.893259
200869717.145.1119314
200990817.432.4158294
2010109816.935.4186389
2011118118.546.9218554
2012116618.953.7220626
2013149418.645.0278672
STEP II
YearEntries1 (%)2 (%)Grade 1 noGrade 2 no
200771238.153.5271381
200885136.050.3306428
200991729.647.3271434
201098030.248.1296471
2011103130.952.8319544
2012100425.741.9258421
2013113424.838.3281434
STEP III
YearEntries1 (%)2 (%)Grade 1 noGrade 2 no
200748338.357.6185278
200855238.056.9210314
200952845.862.9242332
201055441.760.1231333
201163539.757.8252367
201265339.259.6256389
201370636.656.7258400


For example, although the percentage of candidates getting a 2 or higher in STEP I fell from 54 to 45 between 2012 and 2013, the total number of candidates who got a 2 or higher rose by nearly 50, from 626 to 672.

The number of candidates getting a 1 or higher in STEP III is pretty consistent over the last 5 years, even though the number of students sitting it is rising - my guess would be that the 250-ish candidates who get those grades are mostly those who take a place at Cam, and many of those who get a 2 or lower are people who were sitting 2 or 3 papers in order to have multiple stabs at meeting an offer at another uni which is based on a single grade.

Going on the raw data for entries and Cambridge offers, Cam give ~500 maths offers, almost all of which will be dependent on STEP III. 706 students sat STEP III last year, so around 200 were sitting it for a Warwick offer, or as back-up for the unis (Bath, Bristol, UCL etc) which include it as an optional alternative to AEA / A* in FM. Some of those students will (perhaps not always correctly) have been confident that they had met their STEP offer after sitting STEP I or II, so might have taken a rather leisurely approach to the paper.

Also, the STEP I examiners reports have commented (unfavourably, but I think that's a bit harsh) on the 'phenomenon' of increasing numbers of Year 12 students having an early attempt at STEP I. Their data shows that the majority of these candidates don't do very well, not due to lack of ability but because STEP also requires practice and maturity.

There are a few anomalies in terms of the actual grade boundaries - occasionally a question turns out to have a shorter or easier solution than the one that was intended, nearly all the candidates solve it, and effectively the paper is now 20+score/100 rather than score/120. In 2012 there was a misprint in Q1 of STEP III which meant that marks were adjusted/moderated by hand.

What is clear is that you are being compared with your cohort - competing for university offers. You can't compare one year of STEP with another year in the way that you can with A-levels because the purpose of the grade curve is not to ensure equity across cohorts but to fit the right people into the right slots within a cohort. All of the slots will be filled (the universities lose money otherwise), so nobody is being short-changed when the boundaries shift from year to year. If your offer includes "a 1 in any STEP paper" what this really means is something like "No matter how many people sit each STEP, produce one of the best 200 scripts in STEP I, or one of the best 250 scripts in either of STEP II or III."

In the end, though, all you can do is your own best - you can't control whether there are 189 (yay!) or 295 (boo!) people who produce a better set of answers than you in any of the STEPs. We all know that it's not perfect - a different set of questions would have 'sorted' the candidates into a slightly different order, and some would have ended up on the other side of the cut, but the reality is that there are simply more good budding mathematicians than there are places on the top maths courses. The data from Cambridge shows that STEP performance (particularly in III) does correlate very strongly with performance on the Maths tripos - more so than any combination of A-level UMS does with any other subject, which suggests that, however imperfect, STEP is significantly less imperfect than any other method we have for selecting university students.

(And in case it is of comfort, 50% of current Warwick students missed their offer and still got in. Meeting your STEP offer is sufficient, but not necessary, for securing your place at university.)
9
reply
X

Quick Reply

Attached files
Write a reply...
Reply
new posts
Back
to top
Latest
My Feed

See more of what you like on
The Student Room

You can personalise what you see on TSR. Tell us a little about yourself to get started.

Personalise

Should there be a new university admissions system that ditches predicted grades?

No, I think predicted grades should still be used to make offers (626)
33.76%
Yes, I like the idea of applying to uni after I received my grades (PQA) (780)
42.07%
Yes, I like the idea of receiving offers only after I receive my grades (PQO) (365)
19.69%
I think there is a better option than the ones suggested (let us know in the thread!) (83)
4.48%

Watched Threads

View All