omar273 said:
My question is, since i was ranked 2nd in my internal assessment, shouldn't i have gotten the 2nd rank exam mark as my moderated assessment mark, (which is my 97) ??
No.
I know that certain people have explained the moderating process as being like that - but it's just a useful analogy. In reality, it's a little more complicated.
The moderating process implements the following principles:
<ul type="square"><li>The top moderated assessment mark must be equal to the top examination mark.
<li>The mean of the moderated assessment marks must be equal to the mean of the examination marks.
<li>The bottom moderated assessment mark should be as close as possible to the bottom examination mark.
<li>Students should not be disadvantaged by 'anomalous' performances in the examination.</ul>Normally, in a class of only three students, these principles can all be satisfied by simply setting the moderated assessment marks to be the examination marks but reordered according to the assessment ranks.
If that isn't the case, as you have pointed out, the fourth principle must have come into play.
The student ranked third performed worse on the exam (relative to the other two students) than was to be expected on the basis of their assessment mark (relative to the other two students). Including this student's anomalous examination mark in the moderating process would pull down all students not ranked first, so their examination mark will have initially been excluded from consideration.
But, hmmm. It's difficult to tell whether that has actually happened in this case.
Can you give us the marks of the other two students, along with your raw school assessment marks?
Captain pi said:
The moderating process takes the highest exam mark and the mean of your cohort in the HSC exam. The moderating process involves pushing these values through a quadratic transformation (I don't know where the third value comes from)
The third value is the bottom examination mark, or, in certain cases, a value slightly less than the bottom examination mark.
If the skew of the assessment marks is conspicuously different from the skew of the examination marks, the fitting of a quadratic polynomial can produce a non-monotonic curve. For example, the moderated mark associated with the lowest assessment mark may be higher than its immediate neighbours. This case occurs when the turning point of the parabola is between the highest and lowest marks.
In cases such as this the bottom moderated mark is determined by calculation rather than set equal to the bottom examination mark. The appropriate figure is the highest possible mark that is both less than the bottom examination mark and which produces a monotonic increasing curve.
Captain pi said:
(However, I thought that if there were a mark significantly outlying, it was not regarded in the moderating process.)
This is also true - a particular test is applied to determine when this is the case.
The test estimates a bottom examination mark for the group using the principle that the difference between the second lowest and lowest examination marks (in standard deviation units) should not be grossly dissimilar to the difference between the second lowest and lowest assessment marks (in standard deviation units).
In symbols the estimated bottom examination mark is given by:
E<sub>1</sub>(
est) =
E<sub>2</sub> + (
S<sub>E</sub> /
S<sub>x</sub>)(
x<sub>1</sub> -
x<sub>2</sub>)
where
E<sub>1</sub> is the lowest examination mark,
E<sub>2</sub> is the second-lowest examination mark,
x<sub>1</sub> is the lowest assessment mark, and so on.
S<sub>E</sub> is the standard deviation of the examination marks, and
S<sub>x</sub> is the standard deviation of the assessment marks.
If the difference between
E<sub>1</sub> and
E<sub>1</sub>(
est) (the lowest observed examination mark), exceeds a certain value then the result is considered atypically low. This value is set at the equivalent of one standard deviation.
When this condition is satisfied, the lowest examination mark and the lowest assessment mark are removed from the group (these do not necessarily belong to the same student) and the moderation parameters for the group are calculated from the scores of the remaining students. The minimum coursework score is then moderated by extrapolation.
I've emailed one of their liason officers about a similar problem before.