Scaling of the course (2 Viewers)

[lyounamu]

I am completely aware of the fact that scaling of the course is determined by the calibre of the students in general.

However, I have been looking at several scaling reports and came to weird conclusion,

scaling of the course is not determined by the overall calibre of the students undertaking the course...rather, it is determined purely by the quality of the students who counted that particular course into their UAI.

I mean, let's take this as an example. The top UAI from people who took SOR I had 100 UAI. And from this, I also presume that there should be have quite a large number of 99.90+. If we look at the overall calibre of the students taking the course, SOR I should never be capped. Why? Because the top ends of the SOR I students are performing extremely well.

On the other hand, SOR II had top student from its total candidature getting 99.90. With this, SOR II should be capped instead of SOR I which has better top students.

From what I see, only explanation to this is that, scaling of the course is determined purely by the people who have taken the course counted towards their UAI.

Think about it, people who get 100 UAI from SOR I course might have been just forced to do SOR I because of their school. And even if they DID get extremely good mark in SOR I (like 49+), it wouldn't really count for them because the scaled mark that they would get from SOR I is too low.

Well, what do you think?

 

[dux&src]

I am completely aware of the fact that scaling of the course is determined by the calibre of the students in general.

However, I have been looking at several scaling reports and came to weird conclusion,



I mean, let's take this as an example. The top UAI from people who took SOR I had 100 UAI. And from this, I also presume that there should be have quite a large number of 99.90+. If we look at the overall calibre of the students taking the course, SOR I should never be capped. Why? Because the top ends of the SOR I students are performing extremely well.

On the other hand, SOR II had top student from its total candidature getting 99.90. With this, SOR II should be capped instead of SOR I which has better top students.

From what I see, only explanation to this is that, scaling of the course is determined purely by the people who have taken the course counted towards their UAI.

Think about it, people who get 100 UAI from SOR I course might have been just forced to do SOR I because of their school. And even if they DID get extremely good mark in SOR I (like 49+), it wouldn't really count for them because the scaled mark that they would get from SOR I is too low.

Well, what do you think?

wow that's really interesting.

 

[lyounamu]

wow that's really interesting.

Yeah it is.

SOR I had better candidates in general than SOR II. I am quite sure about that. But still, it has a lower scaling...I mean, it is capped while SOR II isn't.

 

[dux&src]

Yeah it is.

SOR I had better candidates in general than SOR II. I am quite sure about that. But still, it has a lower scaling...I mean, it is capped while SOR II isn't.

Then what about the people who are forced to take SOR I like you said. THey might just not take the course seriously. As for people taking SOR II they may have opted for SOR II because they have a genuine interest in it.

Something like that..so maybe SOR II actually have the better candidates...???

 

[lyounamu]

Then what about the people who are forced to take SOR I like you said. THey might just not take the course seriously. As for people taking SOR II they may have opted for SOR II because they have a genuine interest in it.

Something like that..so maybe SOR II actually have the better candidates...???

Well, yeah. But at the top ends, people doing SOR I are more academically superior than SOR II meaning that SOR I has better candidates at the top ends. People doing SOR I may take less care for the subject but that doesn't change the fact that many of them are better than top students from SOR II in terms of UAI.

 

[dux&src]

Well, yeah. But at the top ends, people doing SOR I are more academically superior than SOR II meaning that SOR I has better candidates at the top ends. People doing SOR I may take less care for the subject but that doesn't change the fact that many of them are better than top students from SOR II in terms of UAI.

• the top ends of SOR I > SOR II

but isn't the scaling determined by the overall caliber like you said..
some would say standard, senior science is "capped" but then look at the quality of the top-ends...hmm i dunno now

 

[lyounamu]

• the top ends of SOR I > SOR II

but isn't the scaling is determined by the overall caliber like you said..
some would say standard, senior science is "capped" but then look at the quality of the top-ends...hmm i dunno now

Well, that's what I am basically arguing about now.

I thought that scaling was based on the overall calibre of the students.

I have held on to that belief for a long time but now I am questioning it little bit..

 

[dux&src]

Well, that's what I am basically arguing about now.

I thought that scaling was based on the overall calibre of the students.

I have held on to that belief for a long time but now I am questioning it little bit..

You are questioning it.. on the basis of what?

 

[lyounamu]

You are questioning it.. on the basis of what?

On the basis of the scaling report.

 

[LordPc]

seriously mate, what difference does it make?

just study hard, do the hsc and live your life

 

[lyounamu]

seriously mate, what difference does it make?

just study hard, do the hsc and live your life

lol, i agree. May be I am just being quite pedantic.

 

[Dragonmaster262]

• the top ends of SOR I > SOR II

but isn't the scaling determined by the overall caliber like you said..
some would say standard, senior science is "capped" but then look at the quality of the top-ends...hmm i dunno now

What do you mean by 'capped'?

 

[Timothy.Siu]

What do you mean by 'capped'?

capped as in the highest scaled mark? i dunno

 

[Lazarus]

I mean, let's take this as an example. The top UAI from people who took SOR I had 100 UAI. And from this, I also presume that there should be have quite a large number of 99.90+. If we look at the overall calibre of the students taking the course, SOR I should never be capped.

I'm not sure I agree with your logic.

We know that at least one student who took SOR I last year achieved a UAI of 100. You cannot infer anything about the rest of the candidature from the achievements of one student.

This is a difficult issue to analyse because the non-linear transformation used to scale the marks (and impose a cap) distorts the shape of the distribution of marks. We definitely cannot assume that the scaled marks are normally distributed, for example, even if the raw marks had that distribution.

The scaled mean is the only measure of the overall calibre of the candidature, however, and the initial scaled standard deviation is a measure of the variability of calibre within the candidature.

Last year SOR I had a scaled mean of 27.1, showing that the calibre of the candidature was slightly above average, and a final scaled standard deviation of 8.6. The initial scaled standard deviation was roughly 8.2 (scaled maximum minus scaled mean divided by 2.52). Both standard deviation scores show that the calibre of the candidature was much less variable than average, although strictly speaking only the initial standard deviation is a valid measure of variability.

Compare this to SDD which was also capped at 47.7 but which had a scaled mean of 24.3 and an initial scaled standard deviation of 9.3. This is an example of a course where the performance of outliers has offset to a certain extent the negative effect of the overall calibre.

Why? Because the top ends of the SOR I students are performing extremely well.

On the other hand, SOR II had top student from its total candidature getting 99.90. With this, SOR II should be capped instead of SOR I which has better top students.

These assertions seem to suffer from the same problem.

We only have data about one student in each course - the top student. If you decide to pick up Food Technology and you get a UAI of 99 it doesn't mean that everyone else in the course got UAIs of 99 as well. It doesn't say anything about their performance or their academic calibre at all.

From what I see, only explanation to this is that, scaling of the course is determined purely by the people who have taken the course counted towards their UAI.

This obviously can't be true, because whether a course is counted or not depends on the scaled mark received - in other words, scaling happens before courses are counted, not the reverse.

Think about it, people who get 100 UAI from SOR I course might have been just forced to do SOR I because of their school. And even if they DID get extremely good mark in SOR I (like 49+), it wouldn't really count for them because the scaled mark that they would get from SOR I is too low.

This is possible, but note that Table A6 shows that 93% of SOR I students last year completed more than 10 units, so the vast majority of the candidature could have counted other courses instead.

SOR I had better candidates in general than SOR II. I am quite sure about that. But still, it has a lower scaling...I mean, it is capped while SOR II isn't.

It was capped, but it didn't have 'lower scaling' - on average, the marks were increased. It was only at the upper end that the increase might have been diminished.

I thought that scaling was based on the overall calibre of the students.

I have held on to that belief for a long time but now I am questioning it little bit..

I think you should question your assumptions about the candidature first.

I agree that one can be left with a feeling of uneasiness when the capping methodology is scrutinised closely - but (in my opinion) it is not for the reasons you have set out above.

 

[lyounamu]

LOL I got owned.

Thanks James. I understand now. I can see how the scaling is not generally determined by the performance by the top student and etc.

Thanks again for your great knowledge. I understand far much better now.