equal first. (1 Viewer)

[barbernator]

I know this sounds really petty, and pedantic, yet I was told a while ago that for equal first place moderation, they average the 2 best marks for internal marks.

In a hypothetical situation, if say the top marks were 97 and 91, wouldn't it be better for one student to come slightly second by a very small margin 1% or so, and instead of both getting internals of 94, one would get 97 and one 96~ advantaging them both.

So is this method of averaging the 2 top marks true in this case, because it seems like it could be flawed in a way.

 

[deswa1]

YES. I've been thinking about this thing for a while too. I think it would make sense to come slightly second but I want to know what others think.

 

[D94]

I don't think there is enough information out there to give a definitive answer.

It would seem more appropriate if BOS looks at assessment marks to say 5 decimal places or whatever, and determine an outright first. I would also debate the point as to whether 97 would be an outlier mark, seeing as 91, whilst 6 marks away, is in the span of a whole cohort, rather distant for the two highest marks.

 

[horny unicorn]

my friend missed out an a state rank because of this

 

[barbernator]

my friend missed out an a state rank because of this

ahh that sucks! so is it the average of the best 2 marks as far as you know?

 

[horny unicorn]

ahh that sucks! so is it the average of the best 2 marks as far as you know?

it is

she got 50/50 external and was equal first internal
other equal first got <45 in external
her averaged internal meant she didn't get high enough for state
sans equal first would've gotten a perfect mark

 

[barbernator]

hmm ok well if I tie first in a subject, I might have a bit of convincing to do...

 

[brent012]

In a hypothetical situation, if say the top marks were 97 and 91, wouldn't it be better for one student to come slightly second by a very small margin 1% or so, and instead of both getting internals of 94, one would get 97 and one 96~ advantaging them both.

Are the relative gaps remained even with the first rank though? The way first and last rank get the external marks as their internal mark seems as if they might not be included in that calculation?

 

[someth1ng]

It's really quite stupid, the whole internal marks system is heavily flawed. Schools do not need to test based on the syllabus and don't need to mark like HSC etc. Schools are allowed to set their own assessment schedules and can fully decide what to do with it.

 

[cem]

If there are equal firsts over a year's work then they shouldn't be all that far apart on the exams.

Yes they do take an average of the top two externals for the moderated assessment.

Can you suggest a better system than the one we have - where the one common task is used to moderate what the schools have set.

We could go back to what the situation was in the early 70s and before - where only the actual HSC exam counted - no assessments, no ranks, no chance to show what you can do if it isn't an exam...

 

[D94]

Can you suggest a better system than the one we have - where the one common task is used to moderate what the schools have set.

A maths teacher I once had created some software for maths exams, whereby there was a database of questions (I would say something like 3000+ questions) and teachers could select whichever questions necessary for a maths assessment.

Some system could be implemented along the lines of that piece of software, for a whole range of subjects. Teachers could add questions to the system and it would be moderated by some committee, maybe the marking committee. Then it would be up to the goodwill and responsibility of the teachers to execute this appropriately - teachers can choose to give consistently "easy" questions or give some standard, and some discriminator questions.

There could be some courtesy requirement where teachers should give a certain amount of standard questions, some medium level and some hard level questions. Obviously this doesn't have to be followed (as it would be hard to maintain and check up by BOS), but I'd reckon if teachers really wanted students to do well, they'd follow some state-wide consistent requirement.

And there are some subjects which may not seem viable, even Modern History might not seem to work, but most of it comes to the standardising of marking and the goal of objective marking. This could be achieved with a set of sample answers to each question, probably made by whoever submitted the question.

This obviously will be something massive, and not something possible overnight. Just throwing ideas around, but something could be developed along the lines of this.

 

[brent012]

A maths teacher I once had created some software for maths exams, whereby there was a database of questions (I would say something like 3000+ questions) and teachers could select whichever questions necessary for a maths assessment.

Some system could be implemented along the lines of that piece of software, for a whole range of subjects. Teachers could add questions to the system and it would be moderated by some committee, maybe the marking committee. Then it would be up to the goodwill and responsibility of the teachers to execute this appropriately - teachers can choose to give consistently "easy" questions or give some standard, and some discriminator questions.

There could be some courtesy requirement where teachers should give a certain amount of standard questions, some medium level and some hard level questions. Obviously this doesn't have to be followed (as it would be hard to maintain and check up by BOS), but I'd reckon if teachers really wanted students to do well, they'd follow some state-wide consistent requirement.

And there are some subjects which may not seem viable, even Modern History might not seem to work, but most of it comes to the standardising of marking and the goal of objective marking. This could be achieved with a set of sample answers to each question, probably made by whoever submitted the question.

This obviously will be something massive, and not something possible overnight. Just throwing ideas around, but something could be developed along the lines of this.

Theres too many discrepancies between marking with different teachers, only solution is having multiple standardised tests which is easier said than done. But I believe in Victoria there are external half yearlies so it's not impossible.

 

[asianese]

The whole ranking system is slightly flawed, which adds to this problem.

E.g. in maths our head teacher (stats phd) scales all of them with deviations and consideration of averages etc. So the ranking system is 'fairer' but then I don't know how many decimal places they round to, since last year for 2U there were 3 people on 1st. In economics, the teacher (an idiot) does not scale any of the internal marks, just using raw %, which is stupid. So it makes these types of problems bigger!!

 

[Sindivyn]

The whole ranking system is slightly flawed, which adds to this problem.

E.g. in maths our head teacher (stats phd) scales all of them with deviations and consideration of averages etc. So the ranking system is 'fairer' but then I don't know how many decimal places they round to, since last year for 2U there were 3 people on 1st. In economics, the teacher (an idiot) does not scale any of the internal marks, just using raw %, which is stupid. So it makes these types of problems bigger!!

So do the marks remain in decimal form? Cause some people were saying that they round off after each task, which seems unfair.

 

[deswa1]

So do the marks remain in decimal form? Cause some people were saying that they round off after each task, which seems unfair.

Depends on how your school submits marks. The school pretty much has freedom with the marks they send- they could send it off as 90.18 or just as 90.00 if they wanted which is like effective rounding.

 

[Sindivyn]

Depends on how your school submits marks. The school pretty much has freedom with the marks they send- they could send it off as 90.18 or just as 90.00 if they wanted which is like effective rounding.

Ahh, yeah, but what about rounding of individual tasks, for example 13.4 + 12.4 weighted = 25.8 versus 13 + 12 weighted and rounded = 25.

 

[D94]

The whole ranking system is slightly flawed, which adds to this problem.

E.g. in maths our head teacher (stats phd) scales all of them with deviations and consideration of averages etc. So the ranking system is 'fairer' but then I don't know how many decimal places they round to, since last year for 2U there were 3 people on 1st. In economics, the teacher (an idiot) does not scale any of the internal marks, just using raw %, which is stupid. So it makes these types of problems bigger!!

Why is it stupid? Why should you get 80% when you only achieved 72%? Do you know the method used by your head teacher to scale marks?

 

[deswa1]

Ahh, yeah, but what about rounding of individual tasks, for example 13.4 + 12.4 weighted = 25.8 versus 13 + 12 weighted and rounded = 25.

Oh yeah, that's retarded.

 

[Sindivyn]

Oh yeah, that's retarded.

Yeah, I think it's going to cost me in some subjects - if it's true that is.

 

[D94]

Yeah, I think it's going to cost me in some subjects - if it's true that is.

I doubt they round before the final mark. They'd probably put the marks into some excel file and determine the final mark and rank with the inclusion of many decimal places. You could raise the point with your school...not sure if they've submitted marks yet, but it's worth asking.