how does scaling work? (1 Viewer)

[Trebla]

Courses are capped to avoid highly-able students "dumbing down" and excelling in a course with a less-able candidature. Courses are capped on the basis that the English paper is done by most of the candidature, and the maximum variation in that (the Standard Deviation) should be the maximum variation for all courses. In 2003, the maximum scaled standard deviation was +2.52 and so if a course had students with scaled standard deviation above +2.52, then that course was capped to a mark which was SD = 2.52.

Hold on, so does that mean capping only occurs in subjects that offer courses with specific levels?
What I mean is that some subjects don't offer courses that have specific levels while others do.
For example: Mathematics has many levels of courses:
- General Mathematics (lowest)
- Mathematics
- Mathematics Extension 1
- Mathematics Extension 2 (highest)
While say Science doesn't have a level of courses:
- Biology
- Chemistry
- Earth and Environmental Science
- Physics
- Senior Science

So say General Mathematics is supposedly capped since it is the lowest level of Mathematics, but Science subjects are not capped since there are no low or high levels of Science and all able and non-able students are able to do the same content together in any Science course. Is this the case?

 

[Lazarus]

I don't see why the raw marks have to used for scaling. Since aligning is done to supposedly even out the distribution of marks in comparison to previous sucessive years, why not use that mark, since it is more "fair"?

Aligning ensures that the marks awarded are consistent with the performance bands and the standards that have been set for each course. It has the effect of compressing and inflating the distribution of raw marks, which typically spans a range of 0 - 100, to the range of 50 - 100. This means that students who might have had different raw marks could end up with the same aligned mark (and would thus be ranked the same when they shouldn't be). The UAI is all about ranking students, so raw marks are used to ensure that they can be ranked properly.

Plus, I've noticed on the UAC website (http://www.uac.edu.au/pubs/pdf/2004-Table-A3.pdf) that the scaled mark in MOST cases is lower than the HSC mark, but for some courses like Mathematics Extension 1, the 50th and 25th percentiles have scaled marks higher than their HSC mark, but the 99th, 90th and 75th percentiles have their scaled marks lower than their HSC mark. Why is that the case? Shouldn't higher percentiles have higher scaled marks in relation to their HSC mark compared to those in the lower percentiles? Is this where capping comes in?

As Rench said, you can't compare aligned marks between courses. But more importantly, you can't even compare aligned marks with scaled marks. Aligned marks are derived from raw marks, and scaled marks are derived from raw marks, but both derivations use completely different procedures - it can be meaningful to compare aligned marks with raw marks (to examine the effect of the aligning process), and to compare scaled marks with raw marks (to examine the effect of the scaling process), but there is no value in comparing aligned marks with scaled marks.

Hold on, so does that mean capping only occurs in subjects that offer courses with specific levels?

The term "dumbing down" has been abused a little - you're right, the term does imply that there are higher level courses which were more appropriate and could have been taken. In the past, the maximum scaled mark in every course was guaranteed to be 100% whenever the maximum raw mark was 100% (i.e. the Technical Committee on Scaling would never cause a mark of 100% to be lowered). Furthermore, the Board was in the practise of awarding a mark of 100% to the student(s) who topped the course.

As a result, some students who were capable of taking the higher level English courses would in fact choose the lower level English courses - as they're taken by 'less able' students, it was easier for them to top (or come close to the top of) the course, and hence to be guaranteed a scaled mark of (or close to) 100%. This is usually what is referred to as "dumbing down".

However, it's not restricted to subjects where there are graded courses of varying difficulty. A student bent on maximising their UAI could opt for a course taken by a candidature that was far below average (indicated by a low scaled mean) with the hope of ranking first and being guaranteed that elusive 100%.

To avoid all of this exploitation, the Technical Committee decided to allow the maximum scaled mark to vary between courses, linking it to the overall academic ability of the candidature (the scaled mean) and the variation of that ability (the scaled stdev). As a result, courses with less able and/or less variable candidatures have a maximum scaled mark that is lower than 50 - thus removing the advantage that could be obtained by "dumbing down" under the old system.

Hope that makes sense.

(My personal view is that the scheme didn't go far enough - candidatures of low ability/variability now have lower maximum scaled marks, but candidatures of high ability/variability don't have higher maximum scaled marks. They're 'capped' at 50. I believe that if the maximum scaled mark is to be allowed to vary between courses, it should truly be allowed to vary; even above 50.)

 

[Captain pi]

(My personal view is that the scheme didn't go far enough - candidatures of low ability/variability now have lower maximum scaled marks, but candidatures of high ability/variability don't have higher maximum scaled marks. They're 'capped' at 50. I believe that if the maximum scaled mark is to be allowed to vary between courses, it should truly be allowed to vary; even above 50.)

Hear! hear!

 

[~ ReNcH ~]

I believe that if the maximum scaled mark is to be allowed to vary between courses, it should truly be allowed to vary; even above 50.)

Hehe...
A scaled mark of 105/100 for Ext 2 Maths...nice
If courses were "capped" above 50, then people such as Laurie (and sorry for continually using him as an example) may have got an aggregate of 500...which would be kind of strange.

 

[Lazarus]

Why strange?

The only reason the number 500 has any significance is because the maximum possible mark in each unit is 50, and 10 * 50 = 500; making it the maximum possible aggregate. Obviously, with higher maximum scaled marks, the maximum possible aggregate would also be higher (so he still wouldn't be achieving a "perfect" aggregate).

In Victoria, for courses which have particularly high scaled means, the maximum scaled mark is allowed to vary above 50 (to a maximum of 55), and as a result there is no predefined maximum aggregate. But it doesn't matter, as the aggregates don't have any meaning in themselves and are only used to rank students.

 

[~ ReNcH ~]

True.

Although I think introducing such a system would create more confusion amongst students...which might not necessarily be a good thing. But othet than that I guess it'd be ok to allow for scaled marks >50 for subjects with strong candidatures

 

[Trebla]

As a result, some students who were capable of taking the higher level English courses would in fact choose the lower level English courses - as they're taken by 'less able' students, it was easier for them to top (or come close to the top of) the course, and hence to be guaranteed a scaled mark of (or close to) 100%. This is usually what is referred to as "dumbing down".

Nowadays, English Standard and Advanced are somewhat placed on the same scale. So doesn't the "dumbing down" event still happen as well? According to other threads, English Standard is not capped. So if the system is now more fair, what now stops students, capable of higher level English courses, from taking English Standard for reasons other than personal interest [which would have probably been the case in the past anyway]?

 

[Captain pi]

Nowadays, English Standard and Advanced are somewhat placed on the same scale. So doesn't the "dumbing down" event still happen as well? According to other threads, English Standard is not capped. So if the system is now more fair, what now stops students, capable of higher level English courses, from taking English Standard for reasons other than personal interest [which would have probably been the case in the past anyway]?

English (Advanced) and English (Standard) together are not capped. But, if you top English (Standard), it is unlikely that you will receive full marks. This year, the highest scaled mark in English (Standard) was 93.0/100.0: a 'capping' of 7.0 marks.

I think the reason why English (Standard) seems to perform at a lower standard than English (Advanced) is that English (Standard)'s course has a lower level of sophistication which makes it more difficult for students to acquire higher band descriptors. (If a marker reads a response of Away, he is going to be less impressed than reading Skrzynecki, as I believe.)

By the way, I noticed that you are studying 6U of Science and 3U of Mathematics: Bravo!

 

[Captain pi]

True.

Although I think introducing such a system would create more confusion amongst students...which might not necessarily be a good thing. But othet than that I guess it'd be ok to allow for scaled marks >50 for subjects with strong candidatures

I don't think easing confusion amongst students is a priority for the Technical Committee on Scaling.

I personally think that UAIs should not be released to students: only Universities should have access. (I know this sounds a bit rich coming from someone who has made a FOI application.)

 

[~ ReNcH ~]

I don't think easing confusion amongst students is a priority for the Technical Committee on Scaling.

I personally think that UAIs should not be released to students: only Universities should have access. (I know this sounds a bit rich coming from someone who has made a FOI application.)

But if students weren't given access to their UAIs, they wouldn't have the opportunity to change their preferences if they attained a UAI above what they expected...which could prove troublesome.

 

[Trebla]

English (Advanced) and English (Standard) together are not capped. But, if you top English (Standard), it is unlikely that you will receive full marks. This year, the highest scaled mark in English (Standard) was 93.0/100.0: a 'capping' of 7.0 marks.

You're saying that English Standard is capped by 7.0 marks. That is a darn lot. It's very likely that the highest raw mark obtained in Standard wasn't even 100, probably explaining why the scaled mark was not 100.
(For the record, how do you know thats the highest English Standard mark for this year when the HSC exams for this year have not even started yet?)

Also, I've seen other threads saying that English Standard is not capped even when reported as an individual course. Now you've confused me again.

Doesn't capping only happen to raw/scaled marks not aligned marks?

Are English Standard and Advanced aligned differently or is it just the performance of the candidature?

If capping is done to raw/scaled marks then doesn't that mean that English Standard is not capped since its on the same scale as English Advanced?

IF English Standard is not capped then I'm back to where I started:
According to other threads, English Standard is not capped. So if the system is now more fair, what now stops students, capable of higher level English courses, from taking English Standard for reasons other than personal interest [which would have probably been the case in the past anyway]?

By the way, I noticed that you are studying 6U of Science and 3U of Mathematics: Bravo!

I'm in hell... I want to do only 4 unit Science, 4 unit Mathematics and 2 unit English by Year 12, but then I'd be risking it with 10 units....

 

[withoutaface]

A subject is only capped by its candidatures performance in other subjects. If Std English people performed extremely well there could be a 100.

 

[Estel]

Trebla: Standard and Advanced English are scaled as one, so you still need to get the highest raw mark out of the total candidature to get 100 in Standard.
It follows that if you are good enough to get close to doing that you do Advanced.

 

[Trebla]

A subject is only capped by its candidatures performance in other subjects.

How does that work? I thought scaling and capping happen in each specific course and are not in any way affected by the performance across other courses.

Standard and Advanced English are scaled as one, so you still need to get the highest raw mark out of the total candidature to get 100 in Standard.
It follows that if you are good enough to get close to doing that you do Advanced.

That means Standard is NOT capped at all. It's just so happens to be, that the performace of the candidature in Standard is not as good as Advanced. Am I correct?
If so, then after all that, it's back to my original question...
What stops students, capable of higher level English courses, from taking English Standard for reasons other than personal interest?

I think the reason why English (Standard) seems to perform at a lower standard than English (Advanced) is that English (Standard)'s course has a lower level of sophistication which makes it more difficult for students to acquire higher band descriptors. (If a marker reads a response of Away, he is going to be less impressed than reading Skrzynecki, as I believe.)

Lower level sophistication makes no sense, that just implies that Advanced English students could easily take the "dumbing down" option for English only. Also, I know that a lot of schools make English Standard and English Advanced students do exactly the same texts in the same analytic nature for the Area of Study.

 

[Estel]

Standard is not capped, correct.
But coming top of standard doesn't get you a 100.
Getting top of the complete standard and advanced cohort does.

A high level ability student could do Standard English... but there's no margin of error. Unless they come first in the state in both Paper 1 (for Standard that is, and even then it might not be good enough) and 2, they are screwed.
Given that criteria for Standard also involves less emphasis on artistry and flair, no talented student would be wise to pick it.

 

[Lazarus]

Doesn't capping only happen to raw/scaled marks not aligned marks?

Only scaled marks are capped. Raw and aligned marks are never capped.

Nowadays, English Standard and Advanced are somewhat placed on the same scale. So doesn't the "dumbing down" event still happen as well? According to other threads, English Standard is not capped. So if the system is now more fair, what now stops students, capable of higher level English courses, from taking English Standard for reasons other than personal interest [which would have probably been the case in the past anyway]?

It's true that neither course is capped. Placing the two courses on a common scale prevents students from dumbing down (or rather prevents them from achieving any real advantage in terms of marks by dumbing down).

The distribution of marks on Standard Paper 2 is shaped to coincide with the distribution of marks obtained by the Standard students on Paper 1. Similarly, the distribution of marks on Advanced Paper 2 is shaped to coincide with the distribution of marks obtained by the Advanced students on Paper 1. Estel really hit the nail on the head - you essentially have to top both papers to top the course, and this means you're competing against both Standard and Advanced students, irresepective of which course you're actually taking.

But even if you do top the course, an aligned mark of 100 is far from guaranteed. You'll only receive that if you score 100% of the available raw marks.

Are English Standard and Advanced aligned differently or is it just the performance of the candidature?

Yes - the two courses are aligned differently. We don't know why this is the case. We can, however, deduce that the raw band cut-offs for Advanced are lower (and I suspect in some cases substantially lower) than the corresponding cut-offs for Standard.

English (Advanced) and English (Standard) together are not capped. But, if you top English (Standard), it is unlikely that you will receive full marks. This year, the highest scaled mark in English (Standard) was 93.0/100.0: a 'capping' of 7.0 marks.

This is misleading - no capping occurred. The fact that the actual maximum scaled mark was only 93 (i.e. less than 100%) is immaterial. It was not a maximum mark that was artificially imposed on the course. The maximum possible scaled mark was still 100%.

A subject is only capped by its candidatures performance in other subjects.

How does that work? I thought scaling and capping happen in each specific course and are not in any way affected by the performance across other courses.

The maximum possible scaled mark (i.e. the cap) for a course is determined according to its scaled mean and scaled standard deviation.

The scaled mean and scaled standard deviation for each course is calculated from the performance of the candidature taking that course over all of the courses taken by that candidature. In this way, the course cap is linked to the ability of the candidature taking the course.

 

[~ ReNcH ~]

The distribution of marks on Standard Paper 2 is shaped to coincide with the distribution of marks obtained by the Standard students on Paper 1. Similarly, the distribution of marks on Advanced Paper 2 is shaped to coincide with the distribution of marks obtained by the Advanced students on Paper 1.

So suppose hypothetically that the Standard students did just as well as the Advanced students in the AOS paper (in general), could the highest scaled mark obtained by a Standard student quite well be 100?

 

[Captain pi]

English (Advanced) and English (Standard) together are not capped. But, if you top English (Standard), it is unlikely that you will receive full marks. This year, the highest scaled mark in English (Standard) was 93.0/100.0: a 'capping' of 7.0 marks.

This is misleading - no capping occurred. The fact that the actual maximum scaled mark was only 93 (i.e. less than 100%) is immaterial. It was not a maximum mark that was artificially imposed on the course. The maximum possible scaled mark was still 100%.

(Misleading? I said "'capping'" not "capping". How is that misleading? )

 

[Lazarus]

So suppose hypothetically that the Standard students did just as well as the Advanced students in the AOS paper (in general), could the highest scaled mark obtained by a Standard student quite well be 100?

Yep. Then there'd be nothing to distinguish them from the Advanced students.

 

[~ ReNcH ~]

Yep. Then there'd be nothing to distinguish them from the Advanced students.

Aha...
I'm beginning to understand it now