What if someone above you screws up the exam? Does your mark get dragged down
To help make it simpler for those simple-minded: despite all the complex maths & helpful explanations on BoS the answer is YES, subject to a condition.
If your cohort is large enough such that the mark gaps between adjacent ranks (both school and exam) are no more than 1-2 marks, it is good enough to say you do get the Nth exam mark.
Where the gaps are 3-4 or more marks you can still use that "formula" but have to accept an error margin about half the gap.
And just in case anyone is unsure, the Nth mark won't be your HSC mark.
Your HSC mark is the average of your Own exam mark and the Nth exam mark. Hope this helps.
What if someone above you screws up the exam? Does your mark get dragged down
Again based on no large gaps between ranks you only get dragged down a little, maybe half a mark.
If he's 2nd you're 3rd internally, then he screws up drops to 10th exam (& assuming everyone else performs as expected), you'll get the internal 4th rank's exam mark which should not be much lower than 3rd's.
If he's 2nd you're 10th internally, then he screws up drops to 10th exam, you'll get his mark but it's not too different to the real 10th's.
No, this is pointlessly misleading. It doesn't matter how close the mark will end up being - the answer is unconditionally no. There is an entire process that has been long established as the fairest system for determining the assessment mark. Dumbing it down to nth rank gets nth mark subject to conditions does not further educate students on the correct process. We want to explain correct processes. If they are concerned about moderation, and they don't understand it, the advice should be to just focus on their study and not be concerned about processes that will happen whether they understand it or not. Going out of your way in giving confusing conditional euphemistic solutions doesn't better them in any way, so there is really no point in giving it in the first place.
You might disagree, but even with your explanation, the questions will continue to come. It's better to spread correct information than to appease those who don't understand.
I will argue on maths basis not on the long winded process.
Whatever that process involves the bottom line it ends up with is the cohort's Total M (Moderated Marks) equals their Total E (Exam Marks). In maths sense M is essentially a redistribution of E. If you don't agree with this you should look further.
The distribution of individual marks within M is dependent on that of their internal marks I. So, except a small effect due to Quadratic Interpolation, M is a mapped image of I such that it brings Total I to equal Total E (to eliminate school's harsh or generous marking).
If both I and E are evenly gapped they result in Nth rank getting Nth exam mark. I'd like to see if you can find examples showing they don't.
Last edited by A1P; 19 Apr 2017 at 11:33 PM.
You can't postulate a statement saying "nth rank = nth mark", then add conditions to it and still claim that your initial statement is still true.
http://www.boardofstudies.nsw.edu.au...mple-table.gif
Yes it's a redistribution. That's well established since the mean of the exam marks must = the mean of the moderated assessment marks.
The distribution of the marks is a function of the maximum raw mark, minimum raw mark, mean of the raw marks, standard deviation of the raw marks, maximum exam mark, minimum exam mark and mean of the exam marks. -- so not just the internal marks; the exam marks are factored in, and it's not just pinning the highest and lowest exam marks - the exam marks are factored into the coefficients of the quadratic distribution.
In the example linked, it's clearly shown that the nth rank doesn't simply get a redistribution of the nth mark. There are variations, up to 3 marks, which is significant especially in the higher range of moderated marks. 3 marks across 10 units especially in subjects that scale poorly can result in an ATAR difference of ~15-20 aggregate points or 3 ATAR points (90 vs 87).
I can understand why you give this explanation, but it's not a correct representation of the process. It's far too simplified and the complexities of the process aren't just simplified - they are removed completely, which is inappropriate if you claim that this is what is actually done.
Last edited by D94; 20 Apr 2017 at 12:56 AM.
I wrote at beginning "subject to a condition", you chose to view it as adding conditions afterwards.
The example uses a cohort of 6 students, whereas I wrote at beginning if the cohort is large enough (6 is not large).In the example linked, it's clearly shown that the nth rank doesn't simply get a redistribution of the nth mark.
The school mark gap between B/2nd and C/3rd is 3 vs 12 to one above 17 to one below - totally uneven gaps.
Now if you alter B's & C's school marks to 81 & 72 instead so that they are evenly spaced with A's 90 (the condition I said subject to) then project them on the curve, their Moderated marks would be 80 & 71. Very close to 2nd & 3rd exam marks 80 & 72.
My first post may have given the impression that's how it's done, no I meant that's how the process eventually results in for large cohorts. Not spot on, but close enough to be practical while 90% easier for maths-disinclined students to grasp.
...which is exactly what "subject to a condition" means.
What is objectively true is that the nth rank does not receive the nth mark. That is undeniably a fact.
What has been long acknowledged is that most schools are not sufficiently large, and the distribution of raw marks and exam marks tend to differ more than remain similar. This is the exact reason why the linear model was rejected and the new quadratic model was implemented.
Now you're just describing a scenario that doesn't typically occur. You're just addressing the 1%, not the other 99%, yet you are trying to make it seem this is the usual case by virtue of creating this thread.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Bookmarks