Leaving Cert 2021 Grade inflation

not at the teacher/school level but centrally they did when they issued the final grades , My kids school was furious last year because the students were marked down in a particular subject where they outperform any kind of national average by a large margin.

A grade they would have gotten if they sat the LC itself?

It is clear now from the data that teachers have given out much higher grades thatn

This is a stark reminder of the differences in points between 2010 and now:

what happens if somebody from 2010-2020 applies to do a course? What is a very high mark relative to their cohort will be deemed insufficient compared to this year’s cohort. Furthermore, for those who don’t get their desired course this year: they will most likely reapply next year. If next year’s students follow a more normal curve, does that mean that they will be competing with the higher scoring cohort from this year? Does that mean that inevitably the high-scoring courses next year will be occupied by the 2021 cohort?

Using the figures from the examiner, and noting that 57,952 students sat the traditional LC:

2.32% received 625 points

(1% and 0.2%* in 2020 and 2010 respectively)

*represents maximum achievable points, excluding maths bonus points as N/A in 2010

3.43% received 600-625 points

20.94% received 500-599 points

27.94% received 400-499 points

(the median will lie between 400-450 points)

(the median in 2020 lay between 350-400 points)

(the median in 2010 lay between 300-350)

22.2% received between 300-399 points

13.02% received between 200-299 points

6.25% received between 100-199 points

3.91% received less than 100 points

I think there's two questions there:

(1) Should we be using numerical point grades rather than the blunter instrument of grade ranges (H1, H2 etc)? I think this might improve the present situation of large numbers of students clustering with the same number of points. I asked the same question yesterday as I don't see what the purpose of grade ranges is. So, instead of having whatever number on 625 points, there might be a much smaller number with 625, and then others with 624, 623 etc. I presume the dept of ed has this information, and could answer that question very quickly.

(2) So, if you accept grade points, should they be normalised to something like a percentile? This would make it fairer to transfer grades between years. But, I'm not sure that's a compelling reason on its own. The vast majority of "competition" between students is within the same year's group and percentiles have their own issues as well.

For those who argue against the points system and who advocate aptitude exams or simply counting the relevant subjects, there is a very interesting study (albeit quite old):

http://www.tara.tcd.ie/bitstream/handle/2262/3992/jssisiVolXXIVPartI_231266.pdf;jsessionid=3800B76C45B09CB85A73FD61F04DC323?sequence=1

Additional subject level minimum standards can be required though independent of points. In this case, the maths requirements are not high. If that's because the maths content of the course is not that high, then that's reasonable. If it is deliberately being set low to attract students then that would be a problem

I would not presume to dismiss the calibre or difficulty of a course, particularly based on a title. I imagine that there will be a significant amount of difficult coursework in anything that involves software design. I know somebody in France who did software design for gaming and it was bloody tough! This is probably why the attrition rate is so high. Material science in UL is another notoriously challenging subject.

I agree that having certain subjects do help for certain courses. However, there are students who successfully complete medicine who have never studied LC biology. Similarly, I know somebody who studied German in college and didn’t study it for the LC.

If a person achieves higher points, it does demonstrate a certain degree of motivation, tenacity, capacity to work incrementally in an academic setting. These skills can help the student, if motivated and willing to slog it, if they wish to pursue the course in the absence of the LC course.

It is a tough point to argue against…

Interesting study cited in the IT that reinforces your point, Plodder

https://www.irishtimes.com/news/education/drop-out-rates-in-some-third-level-stem-courses-hitting-80-1.4522466

In France, there is huge consternation. Anybody can start a degree in medicine, regardless of their points in the LC (BAC). They do the matriculation after the first year university. Only the top 20% of the class make it through to second year. A person can repeat the year twice. After this, they must leave. I have met some very bitter people who felt that 2 critical years of their life were wasted by the broken system in France. They leave with nothing to show for their 2 years study after “flunking” out. There are talks about reforming the system in France, to complete the matriculation prior to entry to university. Similarly, to study engineering or business in a prestigious institute, one has to attend an école préparatoire post-LC (BAC). This involves 2 years of very rigorous study. They all write an entrance exam. Based on the ranking, the student may or may not be selected for their desired course. Basically, matriculation is kicked down the line. A sad but true reality is that there are only so many jobs for doctors, engineers, quants, solicitors. A person with a medical degree has to complete their intern year within 2 years to practice clinically. Imagine the outrage if so many medical students were trained, that it was impossible for all of them to get the clinical accreditation to practice medicine. Similarly, LC points do correlate with completion of a course. At some point, there will be a matriculation of some description. And the French system incurs significant more economic cost and runs the risk of wasting the students’ time.

I can see how giving points at a percentage level added to the present grade system would lead to an increase in appeals. Then students who are close to the next higher grade have an incentive to do that. But, if we replaced the grades with just percentage points then I don't see how that would lead to increased appeals. Also, to be really certain that clustering is not a problem, you would need to see the distribution of points allocated in the current system, and then the distribution of points if they were allocated based on the percentage scores directly. Is that data published anywhere?

The problem with percentages for individual subjects is that there is inter-examiner variability. Even a 0.1% difference may make the difference. It would lead to an intense scrutiny of every mark awarded (or not awarded!).

However, my suggestion would be to take the persons LC marks (e.g. if somebody received 520 points in 2010, look at where that falls on the cumulative ranking relative to that year group- 94.2% centile), and award based on centiles. Taking another example, if 2.32% received maximum marks this year, this would be comparable to the top 2.3% who when the tables are examined achieved 555 points and above or 585 points and above in 2010 or 2017 respectively. Then, if you have 2 people from the same year who are tied on centiles, then rather than random based selection, it is awarded to the person from that year who has higher points.

It is not a perfect solution, but I am already imagining the outrage next year when somebody who achieves high marks relative to the rest of the exam cohort loses out to those people from this year who have artificially high points. There will be many people who missed out on their dream courses this year due to random selection looking to change courses next year. There needs to be a fairness to the whole process so that we are not comparing apples and oranges

If they try to solve it by maintaining the current levels of the LC grades in subsequent years, everybody prior to 2021 will be at a major disadvantage when applying to the CAO