State Rankings (1 Viewer)

[Captain pi]

On what is the rank in state based?

Comprehensively.

If a mark of 75/84 translates to a 96% in Extension 1 Maths, then how could someone with an HSC mark of 100 not be in the Top Ten?

 

[Lazarus]

The rankings are determined from raw HSC marks calculated to one decimal place.

A student with a reported HSC mark of 100 (but with a raw mark less than 100%) would miss out on the top 10 if there were 10 students ranked above that student according to raw marks.

 

[Captain pi]

But aren't raw marks in integers?

That is, for Mathematics Extension 1, the top ten marks would be [84, 83, 82, 81, 80, 79, 78, 77, 76, 75]. Now, there would be equal places of course; according to the Top Acheivers List implication of equal places, the marks for each ranking would be:

1st 84/84
2nd 83
4th 82
6th 81
7th 80
8th 79

And, yet, I know a guy who (he says, and I think he's truthful) received a mark of 100/100 (exam) and yet did not get into the top ten. And we know from FOI applications that a mark of 75/84 was 48/50.

I have to go, but hopefully you can see my point.

 

[Lazarus]

But aren't raw marks in integers?

I said "raw HSC mark", not "raw examination mark".

The raw HSC mark is the average of the total weighted mark and the initial moderated assessment mark. The rankings are based on this combined mark, which is not necessarily an integer.

 

[velox]

damn, i thought that state placings were based on the HSC exam only. So that means, you need to be First in your course at your school to get a placing, (unless your school is elite) right?

 

[Lazarus]

That's pretty much right.

 

[Plebeian]

damn, i thought that state placings were based on the HSC exam only. So that means, you need to be First in your course at your school to get a placing, (unless your school is elite) right?

You would think that on the whole, anyone who had a chance at topping the HSC exam would probably be ranked first in his/her school anyway.

 

[eViLnUt]

damn, i thought that state placings were based on the HSC exam only. So that means, you need to be First in your course at your school to get a placing, (unless your school is elite) right?

and what if ur school is a selective school?

 

[velox]

No not really, look at grammar in the classical languages

On a more serious note, ppl can study a lot before the hsc exam.

 

[Slidey]

I said "raw HSC mark", not "raw examination mark".

The raw HSC mark is the average of the total weighted mark and the initial moderated assessment mark. The rankings are based on this combined mark, which is not necessarily an integer.

That is interesting. Something for me to keep in mind.

So as long as I am first, even if first at, say, 20%, it doesn't matter because it is the moderated internal mark...?

Suppose I come uniquely first in a subject, score 90% in that subject - the raw HSC mark would then be 90?

 

[Lazarus]

Yes.
[comment]10 chars[/comment]

 

[Captain pi]

I said "raw HSC mark", not "raw examination mark".

The raw HSC mark is the average of the total weighted mark and the initial moderated assessment mark. The rankings are based on this combined mark, which is not necessarily an integer.

Since his mark was 100/100, I presume he was ranked first and his examination mark was 100/100 and his examination mark was 100/100. Now this may mean that his HSC mark was 99.5.

Now, 75/84 implies 48/50 and, hence, maximally, 96.9/100. This means that the E4 band linear tranformation from TWMs (x) to Aligned marks (y) is, to increase the spread at the top:

y = (3.1/9.0)(x – 75) + 96.9 (1)

Or, if 75/84 implies 95.0/100

y = (5.0/9.0)(x – 75) + 95.0 (2)

(1)

Let y = 99.5 (maximum for Y = 100; Y:Z*)

x > 82

(2)

let y = 99.5

x > 83

How does this work? We know that Ivan Guo received a mark of 100 but was 'only' ranked fourth which would mean that he received a raw examination mark of 82 (assuming a consistent result with the assessment).

Yours,

pi.

 

[Lazarus]

This is intended as a response to Captain pi. I don't expect others to necessarily be able to follow this post.

All comments refer to the Mathematics Extension 1 candidature of 2004.

Assumptions
<ul type="square"><li>A1: The student in question received an aligned examination mark of 100/100.
<li>A2: The student in question was ranked absolute first in his internal assessments.</ul>
Inferences
<ul type="square"><li>I1: The student received an unrounded aligned HSC mark of at least 99.5/100. (From A1, A2.)</ul>
Facts
<ul type="square"><li>F1: A total weighted mark of 75/84 equated to an aligned examination mark of 48/50.
<li>F2: Ivan Guo, a top achiever ranked equal fourth, received an aligned HSC mark of 100/100.
<li>F3: The rankings of the other top achievers are known. (Source)
<li>F4: The student in question was not a top achiever.
<li>F5: The maximum raw examination mark was 84/84.</ul>
Deductions
<ul type="square"><li>D1: Ivan Guo's raw HSC mark was strictly in the range 82.6 ≤ x ≤ 83.8. (From (1), (2), F2, F3.)
<li>D2: A raw HSC mark of 83.5 is sufficient (but not necessary) to be classed as a top achiever. (From F3, F5.)
<li>D3: The student in question's raw HSC mark was strictly in the range 82.6 ≤ x ≤ 83.4. (From I1, D1, F3, F4.)
<li>D4: The top achiever cut-off was strictly in the range 82.7 ≤ x ≤ 83.5. (From D3, F4, F3.)</ul>
All of these deductions are consistent; there are no contradictions.

There are insufficient data to determine whether or not the student in question should have been classed as a top achiever.

If a contradiction were to be found, I would question the validity of A2.

 

[Captain pi]

83.5

How does one achieve a raw HSC mark of 83.5?

Of course, 83.5 could be achieved from a raw exam mark of 84 (or 83) and an initial moderated assessment mark of 83 (or 84). But if the student were ranked absolute first, this would be impossible.

How could a student, who was ranked first in the assessment, and achieved the highest mark in the exam for his centre, receive a non-integer exam mark or assessment mark?

 

[Lazarus]

F6: If a student who was ranked first internally also achieved the highest examination mark for their centre, and there were no appeals on the basis of illness or misadventure, their raw HSC mark would be an integer.

D5: The student achieved a raw HSC mark of 83/84. (From D3, F6.)
D6: The top achiever cut-off was strictly in the range 83.1 ≤ x ≤ 83.5. (From D4, D5, F4.)

If any of these deductions are incorrect, at least one of the assumptions must be incorrect.

 

[Captain pi]

(Thanks for spending time on this Lazarus; I'm not suggesting anything's amiss here, I'm just sophisticating¹ my knowledge of moderating and aligning.)

How would a mark of 83.1 or a mark of 83.4 or 83.8 be achieved?

Is there a more complicated moderating algorithm of which I am not aware? Does moderating take into account anything but: (a) the raw examination mark of the given student; and, (b) the corresponding raw examination mark to that student's rank.?

I don't understand how raw HSC marks can be anything but means of integers.

¹(what a word!)

 

[Lazarus]

Moderated assessment marks are rarely integers. This usually only occurs in the cases of the top and bottom students.

Does moderating take into account anything but: (a) the raw examination mark of the given student; and, (b) the corresponding raw examination mark to that student's rank.?

This is not how the moderating algorithm works - though it provides a rough approximation.

The whole of the distribution of examination marks obtained by the class is taken into account.

Students are assigned marks in such a way that the top moderated assessment mark is equal to the top examination mark and the mean of the moderated assessment marks is equal to the mean of the examination marks. In most cases, the bottom moderated assessment mark will also be equal to the bottom examination mark.

The effect of the moderating procedure (which is a quadratic transformation) on other students is usually to push them away from the mean.

Experiment: http://www.boredofstudies.org/moderate.php