Can you comment on two important, relevant issues not addressed in your analysis: the CAO, and in particular the maths bonus points, that have disproportionately affected LC student behaviours; and the issue of curriculum squeeze at senior cycle, whereby due to the expansion of well-being components, teaching time allocated to Maths will reduce from next year (and let's not forget that the current Project Maths spec is universally considered too big already).
Finally, are you advocating that assessment should be by terminal examination only (the status quo) or can you viably see any alternative models?
Education policy is far from my area, so I can't say I've read or thought about any of these issues in detail. But my initial reaction is:
1) The bonus points incentivise the intellectual level of the material at Higher Level to be reduced, and for students who are barely understanding to hold on, if at all possible. It's somewhat of an empirical question how large these effects are compared to the benefit of having a greater number of students enrolling in HL. My instinct based on this post would be that the benefits of the latter are not worth the costs of the former.
2) We were going to include some stuff about the total number of contact hours likely decreasing, but I couldn't find statistics on what the average level currently is, and how many classrooms would be affected by a drop in the minimum class time (which, as I understand it, is the change being mooted). There is a wide range internationally in how long students spend in school per day, and it never seems to correlate with much, which made me feel that unpacking the relationship between class time and student outcomes would require a post of its own.
3) I'm not sure what the problem is exactly with assessing based on written in-person exams for maths. A strong student is able to solve any problem on the current curriculum within a few minutes; the idea that 5 hours isn't enough time to fully capture the nuances of someone's mathematical thinking about such straightforward problems seems a bit far-fetched. I do sympathise with students who are ill or just have a bad day, and it's obviously very stressful, so perhaps there could be a system which averages across a greater number of exams. But I haven't thought that much about it.
I agree with much of what Sam writes. I would only add on (3) that I think the space of possibilities has narrowed quite substantially in the last few years. While many countries have had bad experiences with non-exam assessments for maths (I am especially familiar with England and New Zealand), other countries and also the International Baccalaureate have seemingly made it work.
But these assessments tend to be completed over a rather long period of time (weeks, or perhaps longer, of take-home work), and the problems are relatively open-ended relative to usual homework problems, demanding some amount of originality and personal thought from the kids who are examined. The problem has always been how to ensure that the kids are completing these assessments themselves, and not copying friends / older siblings / the internet / etc. But that seems to have historically been a solvable problem.
AI renders this problem entirely non-solvable: any student given any problem, no matter how open-ended or novel (so long as it is in theory solvable with their mathematical knowledge), can feed it into ChatGPT and get an entirely bespoke solution provided for them with zero effort. Minor human-level tweaking and rewriting can leave this invisible to e.g. Pangram.
This is exactly what led the previous Education Minister to change tack: Foley initially seemed to propose longer-term, potentially take-home assessments. But in 2023 she made it clear the non-exam assessments would be single-day and done under controlled conditions, precisely because of AI.
When you're assessing students on a single day and under controlled conditions, you are essentially holding an exam, whether you call it 'exam' or not. The only question is whether it's a rigourous exam or not. 'Investigating with Mathematics' is the latter: it will be just as stressful for students as current exams, without any of the rigour or reliability.
Important advocacy. Well done both.
Can you comment on two important, relevant issues not addressed in your analysis: the CAO, and in particular the maths bonus points, that have disproportionately affected LC student behaviours; and the issue of curriculum squeeze at senior cycle, whereby due to the expansion of well-being components, teaching time allocated to Maths will reduce from next year (and let's not forget that the current Project Maths spec is universally considered too big already).
Finally, are you advocating that assessment should be by terminal examination only (the status quo) or can you viably see any alternative models?
Education policy is far from my area, so I can't say I've read or thought about any of these issues in detail. But my initial reaction is:
1) The bonus points incentivise the intellectual level of the material at Higher Level to be reduced, and for students who are barely understanding to hold on, if at all possible. It's somewhat of an empirical question how large these effects are compared to the benefit of having a greater number of students enrolling in HL. My instinct based on this post would be that the benefits of the latter are not worth the costs of the former.
2) We were going to include some stuff about the total number of contact hours likely decreasing, but I couldn't find statistics on what the average level currently is, and how many classrooms would be affected by a drop in the minimum class time (which, as I understand it, is the change being mooted). There is a wide range internationally in how long students spend in school per day, and it never seems to correlate with much, which made me feel that unpacking the relationship between class time and student outcomes would require a post of its own.
3) I'm not sure what the problem is exactly with assessing based on written in-person exams for maths. A strong student is able to solve any problem on the current curriculum within a few minutes; the idea that 5 hours isn't enough time to fully capture the nuances of someone's mathematical thinking about such straightforward problems seems a bit far-fetched. I do sympathise with students who are ill or just have a bad day, and it's obviously very stressful, so perhaps there could be a system which averages across a greater number of exams. But I haven't thought that much about it.
I agree with much of what Sam writes. I would only add on (3) that I think the space of possibilities has narrowed quite substantially in the last few years. While many countries have had bad experiences with non-exam assessments for maths (I am especially familiar with England and New Zealand), other countries and also the International Baccalaureate have seemingly made it work.
But these assessments tend to be completed over a rather long period of time (weeks, or perhaps longer, of take-home work), and the problems are relatively open-ended relative to usual homework problems, demanding some amount of originality and personal thought from the kids who are examined. The problem has always been how to ensure that the kids are completing these assessments themselves, and not copying friends / older siblings / the internet / etc. But that seems to have historically been a solvable problem.
AI renders this problem entirely non-solvable: any student given any problem, no matter how open-ended or novel (so long as it is in theory solvable with their mathematical knowledge), can feed it into ChatGPT and get an entirely bespoke solution provided for them with zero effort. Minor human-level tweaking and rewriting can leave this invisible to e.g. Pangram.
This is exactly what led the previous Education Minister to change tack: Foley initially seemed to propose longer-term, potentially take-home assessments. But in 2023 she made it clear the non-exam assessments would be single-day and done under controlled conditions, precisely because of AI.
When you're assessing students on a single day and under controlled conditions, you are essentially holding an exam, whether you call it 'exam' or not. The only question is whether it's a rigourous exam or not. 'Investigating with Mathematics' is the latter: it will be just as stressful for students as current exams, without any of the rigour or reliability.
This is forensic and damning. The same fault-lines have been evident in English for several years. It's a critical moment.
https://juliangirdham.substack.com/p/extra-irish-education-sunk-cost-fallacy
The Philistine Ascendency holds nothing sacred.