Irish Independent - Opinion Piece 08.09.2020
Professor Damian Murchan
Already the debate has moved down the road to CAO offers. After all the drama of calculated grades, the results are in and students are again in wait-mode.'
Many students, parents and teachers are happy. Some are not. The Department of Education agreed largely with grades assigned by teachers - four out of every five were accepted. We knew that a week ago, and this took some of the noise out of the system on results day. On average, the class of 2020 outperformed the class of 2019 and previous years. Some will call this grade inflation on a scale not seen since the 1990s.
Others will see it as compensation for hardships endured by students over the past months. CAO applicants who are using grades from previous years to access college may see it as unfair.
The increases came despite application of standardisation. This was inevitable since pressure mounted on Marlborough Street to scrap so-called school profiling, one element in the algorithm built to generate calculated grades. The key tool to counter grade inflation was gone.
Not all grades went untouched - 4pc of grades were increased, 17pc went down, indicating some standardisation at work. This will be interrogated in the coming period, though not as closely as if 25pc or 39pc of grades were reduced, as happened in Scotland and England. Amid calls for transparency about the algorithm, it is worth remembering this isn't the first time that Leaving Cert grades have been standardised. It happens every year.
Generally, there is great similarity in the distribution of grades within each subject across years. Just ask students. They examine these historical patterns to gauge the odds for achieving high grades in specific subjects. This similarity doesn't happen by chance. In Ireland, as elsewhere, there is an expectation for general stability in national examination grades from year to year.
Given the high stakes associated with the Leaving Cert, students, parents and teachers don't like unwelcome surprises on results day. Consequently, examination agencies and ministers don't like surprises either. So, we generally see the same distribution of grades in a subject each year.
Such stability in grade distributions is engineered also because the exam changes each year. Questions are not tried out in advance, so we don't know how appropriate they are. Nor do we know how easy or difficult they will be. Students could be lucky because the exam is easy. Or unlucky. Students' and teachers' belief that an exam was "harder" may be well-founded and this fluctuation in difficulty is a challenge confronting examination agencies in many countries. How the problem is managed varies.
Officials typically try to maintain the overall distribution of grades - for example, the same proportions of students receiving H1, H2, H3, etc. This is based on the assumption the characteristics of the cohort (all Leaving Cert students) don't change significantly from year to year. In England, the grade boundary at which a grade (such as A or B) is assigned is allowed to vary from subject to subject and from year to year. The devil is in the algorithmic detail, but the distribution of As, Bs, Cs, etc, remains constant.
In the Leaving Cert, standardisation occurs at the point where scripts and other work are being marked by examiners. In this real-time form of standardisation, consistency in the distribution of grades is achieved by adjusting the marking scheme used by examiners. Grade boundaries are fixed in relation to percentages of marks achieved by students on the exam. Every year, a student needs 90-100pc to get a H1/O1 grade; 80-89pc to receive H2/O2 and so on.
Unless some adjustment is applied, it is improbable the same proportions of students will fall neatly into each grade band annually, especially given the variation in exam questions from year to year.
This type of standardisation involves monitoring marking as it unfolds, especially in the early stages. Examiners attend marking conferences immediately after the exams. They discuss the proposed marking scheme and undertake some supervised grading of scripts, helping to ensure consistency in marking across examiners.
That process also yields early indication of likely trends in grades and allows for adjustment of the marking schemes in each subject. If needed, this recalibration can continue in light of analysis of further samples of grading collected from examiners over the subsequent weeks. Surprises are kept to a minimum. The distributions remain the same.
What the recent experience with calculated grades and standardisation has revealed is not the need to consider if the Leaving Cert should be standardised. Rather, how best this should be done. This is a debate undertaken preferably away from the pressurised context of ensuring fairness for one group of students who have had a very tough time.
Prof. Damian Murchan is Head of the School of Education in Trinity College Dublin