HSC English - Identical Scaling
Are English Studies Exam, English Standard and English Advanced really scaled exactly the same?
English Studies Exam, English Standard and English Advanced scale the same in the HSC.
I’m regularly asked about this from schools and teachers. It turns out to be complicated. Not because of the truth of it, but because there are so many articles on the interwebs that say otherwise. So there are plenty of sources to show otherwise and drawing incorrect inferences from true data is very easy.
Also, UAC get this a bit wrong, which is most frustrating of all. But I’ll get to that later.
Here’s what the 2022 UAC Scaling Report says about it.
“As NESA places English Studies, English Standard and English Advanced raw marks on a common scale, these courses are combined and scaled as a single course but are reported as separate courses in order to be consistent with NESA’s reporting practice.”
It’s pretty plain. But, to be fair, that’s referencing raw scores, not aligned HSC scores.
So what does the data point to the same scaling?
Yes. But there is some complexity to it all.
Here is Table A3 in the 2023 Preliminary Report on Scaling, for English subjects.
I’ve seen people use this table to demonstrate how English Advanced scales much better than Standard. They do that using the percentile columns. But to point out that the top 10 percent of students in Advanced achieve higher scaled scores as the top 10% of students in Standard is as useful, I think, as pointing out that water is wet.
What we need to do is look for equivalent HSC scores and then check what their scaled scores are.
This table in the scaling report reports HSC marks out of 50, not out of 100. Scaled scores are also out of 50.
Here’s the same table, with some scores highlighted.
These HSC scores aren’t identical, but two are and the other is close. An HSC score of 38 (76) in Studies and Standard scales to the same scaled score. And a 39 (78) in Advanced scales to a little more than the 38 in the other two.
It gets a little more complicated with some other scores. All three courses have an HSC score of 43.5 (87) reported in the scaling table.
The scaled scores here are a little different, but they’re very close. This is because a single HSC score is worth a range of scaled scores. When the HSC is scaled, it’s done so using raw scores. It’s not until later in the process that scores are compressed to fit into the band system. And the band system doesn’t use decimals. So an HSC score of 43.5 (87) is made up of a range of raw HSC scores that include decimal places.
Here are all the HSC scores from Band 2 up achieved in English Advanced in 2023.
1,316 students in Advanced scored an HSC score of 87.
Here’s Standard:
187 students in Standard achieved an HSC score of 87.
Here’s Studies Exam:
One student in English Studies who sat the exam scored an HSC score of 87.
So across 2 Unit English (excluding EALD), 1,504 students achieved an HSC score of 87 (43.5 in the scaling report). They didn’t all score the exact same scaled scores from that HSC score. But they all achieved within the same range of scaled scores. It’s a small range, but it is a range.
So What?
There is a wide range of articles about which level of English students should study in the HSC. When those articles are ATAR focused, the advice they give to students seems to be that students should study Advanced English, as it will be worth a lot more to their ATAR than Standard. They make this argument by appealing to percentiles.
It’s definitely true that a higher percentage (by many factors) of students in Advanced achieve at Band 6 level than Standard. It’s also true that a lot more Advanced students get higher ATAR contributions from English than Standard students. But the truth of that has no implications as to the equivalency of scaling in the two subjects.
The best level of English for a student requires a few considerations, but which of the English subjects scales better is not one of them. And that, just because they scale the same.
What about UAC’s ATAR Compass?
UAC’s ATAR estimator makes this a little complicated. I put in two sets of results to ATAR Compass. Here’s what I got.
I think I know what causes this and it’s frustrating. It’s only for lower marks in Advanced (that is, marks below the 25th percentile in the cohort). For higher marks, this doesn’t happen.
So, I don’t know. You can believe the scaling report and the data, or you can believe UAC’s ATAR estimator and the tutoring companies, I guess. When I make estimates of student ATARs for schools, I force English Studies Exam, Standard and Advanced onto the same scale and I get better estimates of ATARs from it.
The thing that strikes me personally about this, is how much student self-perception plays into the subject selection and thus the data.
Students don't like the idea of being in studies. If the option among their friend group is standard and advanced and they feel they are on the "top half" of their friends intellectually or whatever, they really don't like doing standard.
So kids feel a need to prove themselves in an education world that has definitely moved away from streaming earlier years by ability. It's their first chance to put themselves in a "class category". And their own data set is absolutely tiny and rarely very objective as English teachers are not rewarded for marking honestly in earlier years. Really the ABCD range is a very relative mix compared to Science or Maths with wrong and right answers. If the whole cohort in a maths exam gets less than 60/100 the teacher's hands are tied. In English it's more tied the other way of "why be cruel?"
And I think all of this flows through into the HSC and HSC exam marking.
We've just gotten back marks and we have a very small cohort so we know them well and we don't have many numbers to compare. And we've seen one of our weakest students who was sick the day she did paper 2 and absent for paper 1. But she got exactly the same mark as our best student who wins awards on other external academic competitions.
I can't shake the sense that the weaker student got marked after the marker read some truly terrible papers, while the stronger student got a different marker with the reverse. Because the weaker student didn't do the short answer component she's been able to use prepared responses for all.
If we take Ext 2 and its volatility, it may be far more typical of how English is marked, but the rest gets "sanitised" by the norming presence of large numbers.