Official Report (Hansard)
Date: 02 March 2011
PDF version of this report (191.31 kb)
Members present for all or part of the proceedings:
Mr Mervyn Storey (Chairperson)
Mr David Hilditch (Deputy Chairperson)
Mr Jonathan Craig
Mr Trevor Lunn
Mr Basil McCrea
Miss Michelle McIlveen
Mr John O’Dowd
Mrs Michelle O’Neill
|Mrs Katrina Godfrey||)||Department of Education|
|Mr Chris Hughes||)|
|Ms Karen McCullough||)|
|Mr Roger McCune||)||Council for the Curriculum, Examinations and Assessment|
The Chairperson of the Committee for Education (Mr Storey):
I welcome Roger McCune from the Council for the Curriculum, Examinations and Assessment. I trust that you are the wiser after that exchange of views.
Mr Roger McCune (Council for the Curriculum, Examinations and Assessment):
There are a lot of similarities in relation to GCSEs and how we make predictions for GCEs using GCSE results.
Mrs Katrina Godfrey (Department of Education):
I am conscious of time, so I will not say anything.
I am head of regulation at CCEA. We regulate all non-vocational qualifications. Five awarding bodies offer GCE and GCSE qualifications across England, Wales and here. We regulate those qualifications in Northern Ireland. We have a common qualification system with England and Wales. Therefore we work closely with their regulators, which are Ofqual and the Department for Children, Education, Lifelong Learning and Skills (DCELLS) in Wales.
Regulators ensure the integrity of the qualification system so that those who use it, particularly employers and higher education providers, can have confidence in the awards of the grades that come out of the system. We call GCSEs and GCEs high stakes qualifications. Therefore it is important that the standard of qualifications are the same, regardless of what awarding body offered it. Therefore it does not matter what awarding body awarded the qualifications to people who are applying to universities, because the grade represents their achievement.
We do that in a number of ways. First, we have strict criteria against which the qualification specifications and the assessment materials are developed. I know that one of the issues that you will want to talk about is the level of the qualification. Accredited qualifications are put into a national framework at a particular level. GCE qualifications are at level 3 in the national framework. The framework has nine levels: an entry level, and levels 1, 2 and 3 up to level 8, which is university PhD. Level 3, or GCE, is the level that we are talking about for university entrants. GCSE is a unique qualification, because of the great spread of the cohort. I will talk a wee bit about that. Over 90% of 16-year-olds will take GCSE qualifications. It is a very broad qualification, and it covers level 1 and level 2. I have been told that it is one of the issues that you want to mention.
I am conscious that I need to get a wee bit technical on grades and levels, so please interrupt me as I go through it. You are right to talk about level 2 and GCSE achievement, because it is important in respect of such skills as communication skills and application of number — as we talk about — or use of mathematics, as they talk about in 4 to 16. In his review of skills, Lord Leitch talked about those as being the jumping off points. He said that level 2 was the jumping off point into employment or into effective living. So, young people’s level of achievement at age 16 is very important, because it has an effect on their progress.
The revised curriculum, which you talked about, focuses on skills. It is, therefore, important that we ensure that, while we attack the problem upstream, there is coherence in the education of pupils between the ages of 14 and 19 by continuing to develop those skills from GCSE through to GCE.
The vast majority of young people begin GCSE courses aged 16, particularly in maths and English. We recognised the importance of that link with skills development to the extent that those who were first taught the revised GCSE maths and English courses last September now have those skills embedded in them. In other words, achievement at level 2 is built into achievement at grade C and above in GCSE. We may want to talk about that, because we worked very closely on it. The revised GCSE started in 2009, but we took an extra year to ensure that we embedded English and maths in the GCSE specifications because of the importance, as far as employers are concerned, of communication skills and the application of numbers.
What are the features of level 2 performance compared to level 1? In other words, what does a grade C and above achieve compared with a grade D and below? We will talk about that issue because, at the end of the day, we do draw lines. We are talking about the average, or mean achievement, at grades C and D. I will cite an evidence statement, which we use to develop skills tests, on just one aspect of communication skills as an example. Level 1 means that a pupil is responding to information and is following explanations and instructions. Level 2 means that the individual is responding to extended information following detailed explanations and multistep instructions. In the application of numbers — in handling data — level l means that a pupil is carrying out calculations to the level of accuracy that they have been given. Level 2 means that the individual is carrying out calculations, showing clearly the methods used and the levels of accuracy, including identifying and correcting any errors. In other words, level 2 is looking at greater depth; it is looking at a wider complexity of information, synthesising and bringing together information and greater independence in learning.
I want to get a wee bit technical now, because I saw the Chairperson nodding when I mentioned grades C and D. I want to talk about how we end up with the grades in GCSE, and I will take English as an example. In GCSE qualifications, one is not looking for failure or whether a pupil can or cannot do something. However, with a vocational qualification, one would want, quite rightly, to look for that. If we give an NVQ to someone who fixes an engine, it is no good saying that he achieved a grade C, and got it a bit right, if the car breaks down 20 miles up the road. With vocational qualifications, it is a case of being right or wrong. With GCSEs, one is looking at a spectrum of achievement from an A* grade, which is very high, to a G grade, which is very low. One is looking at the knowledge, understanding and skills that the young person will develop.
Pupils take two GCSE English papers; a speaking and listening test and a coursework element. The tests are marked out of, let us say, 60 and 60, and the marks are combined to give the score. We then put in the grade boundaries after looking at the examiners’ judgement and the statistical evidence. Guidance is given on GCSEs, and examiners make their judgements based on the grade descriptors for grades A, C and F, which I am happy to read out if we have time.
Then the other grades, such as the grade B, are fitted in based on what we call “statistics”. Let me give you an example. If 72 marks out of 100 are agreed as the A/B boundary in GCSE, and 48 marks are agreed as the C/D boundary, then the B/C boundary would be determined by adding the two marks together and dividing by two, which gives 60 as the arithmetic mean for the B/C boundary.
One could say that the difference between a C and D is one mark. However, we are looking at the broad descriptors of where people sit in the spectrum. So, all achievement for young people, whatever grade they get, is an achievement as far as their education is concerned. We then move to the employers’ side or to higher education where we introduce a competitive element or a utilitarian approach, and where we talk about GCSEs at grade C and above as being at level 2, and grade D and below are at level 1.
That is broadly how the system operates for GCSE. The innovative thing that we have done is to bring the skills dimension into the GCSEs in English and maths. We are marrying two systems: the compensatory system at GSE and the competence-based system of, for example, changing plugs in car mechanics.
We are now going to look at whether we have achieved that, and we will do so in the nature of the questions of whether in GCSE we are testing communication skills in English and application skills in maths. Also, if we are right, if the pupils do a skills test and a GCSE together, we should see those getting a grade C and above in the GCSE getting level 2 in the skills test, and those getting a grade D and below not getting level 2, but getting level 1.
How then, Roger, did we change the descriptor when we moved to define grade A*? The Committee has been concerned about that. Trevor Lunn raised that point, and I will ask him to speak in a moment.
I presume that you are talking about the grade A* in GCSE, rather than the grade A* in GCE, which we have introduced?
I was going to ask you about GCE as well, but let us concentrate on GCSE. How does that affect the overall distinction between level 1 and level 2? Has it had an impact? Level 2 seems to be the minimum, especially with respect to the platform for required skills.
We have had grade A* in GCSE for about eight or nine years. It came about for the same reason as grade A* in GCE has come in, because there are too many grade As — if you want to think about it in that way. We are trying to discriminate within the grade A. The A*/A boundary is a statistical boundary. It is a complicated logarithm. I explained that the B/C boundary is calculated by taking the mean of the A/B and C/D boundaries. In the case of the A*/A boundary, we move half way between the A and the B to get the A* and there is a wee bit of manoeuvring in relation to examiner judgement within that. Basically, grade A* covers between 4% and 7% of the overall cohort and around one third of the grade A cohort. The grade A is around 20% or 25% of the overall cohort.
What was the rationale for it? Was it the success or otherwise of the grade A* in GCSE? Why did you feel that the grade A* was so useful and beneficial that you brought it into GCE?
In A-level and GCE there were two reasons why the grade A* was developed. First, we wanted to increase the rigour of the assessment in GCSE and we have done that by testing GCE in four units: two AS units and two A2 units.
We wanted to introduce what is called a greater stretching challenge within the A2 unit. So, the nature of the questions, the stretching challenge, was increased.
The other, more utilitarian reason, was that, as you know, many universities were saying that, for their high-demand courses — for example, for medicine at Queen’s — it was proving very difficult to use GCE on its own, and that an A* would therefore be helpful. The uptake on that in universities has been interesting. Cambridge has used it since last year, and Oxford is now going to use it this year for maths and science. Queen’s is still not sure whether it should use it for admissions to medicine courses, and that is fair enough, because one could argue that there is more to medicine, if one thinks about the mastery of the subject, whereas maths and science may well be treated differently.
Those were the two reasons that it was brought in. Now, with GCE, the higher-performing element of the A grade is being awarded an A*. I do not know whether that is completely clear.
I am interested. The reason I raised it in the first place was not on my own initiative; it was because headmasters kept asking me and were making the point. We have seen it in investigations carried out recently into schools that perform well in disadvantaged areas, which asked the same question. They all reckon that, if you look at A* to D grades instead of the A* to C grades, there is roughly an extra 15% who would come into level 2 if that included a D grade. You told us that the difference between a C and a D could be as little as one mark.
It worries me, and it worries those headmasters, because some pupils with D grades have made monumental strides since their transfer. Headmasters would like to think that a D grade in maths and English should not exclude a person, in an employer’s eyes, from the list of likely participants in the workplace until they learn some better skills. That worries people. Their view is that a lot of the children who achieve D grades are reasonably high achievers and are perfectly competent to enter the workplace. They may not be experts on Shakespeare, but they can certainly read and write competently. That is the worry.
It is a very fair point. The point that I was making was that we are working with the Department at the entry level in relation to special needs pupils to give better recognition to the achievement of those with special needs. All achievement for young people is important, and schools should celebrate it, as should the young people and their parents. The cut-off point is at grade C. There is no going back to the CSE days, when a grade 1 CSE was the equivalent of a C in the old O-level. There has been that cut-off point of at grade C and above in GCSE.
We have spoken to employers about the most use they have for qualifications. It would not surprise you to know that it is for shortlisting. In broad terms, for level 1 and level 2, we are looking at the two issues, if you want to take in the big picture of the levels, as against what I had to admit to you — that at the end of the day it is about one mark. What is more complicated is that, if you take those elements that make up the one mark, there could be people with 112 points, who are getting a D, who are performing better in some elements of English than those with 113 points. In the big picture, the level 2 qualification looks much more at the greater independence and greater detail for the learner. In broad terms, level 1 and level 2 make sense, but there are issues. It is just the nature of the grading system for GCSE.
Roger, you were saying earlier that the introduction of the A* grade was driven by universities.
Universities were one aspect.
At the other end of the spectrum, to which Trevor referred, are workforce requirements. I know there is an issue of balance in this matter. However, I am not being biased when I say this, but we can all remember the presentation that we received from two of Northern Ireland’s largest employers when we went to a school in Ballymena. They were not looking for A*s, but for young people to come to them with good basic maths, English, and so on. How are we constructing the system so that it really reflects industry’s needs? For example, in my constituency, Wrightbus is a massive employer. To be honest, it needs to employ a range of people. How do we get the balance right? I worry that the current structure is too restrictive.
From our perspective, as regards attainment levels in literacy and numeracy, employers tell us all the time that they attach most value to level 2, which is the point that Roger made. Colleagues in the Department for Employment and Learning say that it is level 2. In fact, its skills strategy will focus, at the very least, on increasing the number of the people in the workforce who are qualified to level 2. It is the added complexity of knowledge and skills application that makes up level 2. Business consistently tells us that, too. As I am sure that you heard from Wrightbus and others, they attach value to that level of complexity. For them, when young people achieve at that level, they are able to take their place in the world of work. As you say, there may be areas where they need much higher qualifications. There may be areas where they need specific qualifications. However, what employers are telling us is that, as a minimum, level 2 in literacy/English and numeracy/maths, and particularly the skills of applying mathematical knowledge, is most important to them.
The key thing for us is that — and Roger is quite right — in every school, there will, inevitably, be the story of one pupil who was perhaps one mark short. I suppose that that is a product of any examination system in which grade boundaries are set. The key point, and that which Roger emphasised, is that for the vast majority of children who achieve a grade C, you will see a higher level of competence than you will see from the vast majority of children who achieve a grade D. That is the key point. That is why our focus has been on supporting many more of those children, particularly those who Trevor mentioned, who are on the borderline, so that they can leave with a full level 2 qualification. That is what the economy demands. That is the key point. In any system — in the same way as at any grade boundary — there will, inevitably, be one or two pupils who slip to one side or the other.
One other point that is, perhaps, worth mentioning — I hope that I am correct on this, Roger — is that the introduction of the A* grade did not actually affect boundaries for other grades. It did not drop them down. The A* grade was inside the A boundary. It did not have a knock-on impact on other grades. Perhaps, that is also a key point.
I must confess that when Trevor started on his crusade I was not particularly convinced. Again, this will fill him with concern, but, increasingly, I am on the same page as him on the matter. There can be a difference of one mark only. We talk about the “winner takes all” society in which we live. We consign people with grade Ds to go left while people with grades A-C go right.
It is a bit like the transfer test.
Mr B McCrea:
It is, Trevor. I am putting exactly the same argument. It did not occur to me that there could be a difference of one mark only. You said that in certain elements, someone’s performance might be better than a grade D. If a grade D is considered acceptable — I am sorry that I might have missed what you said about levels 1, 2 and 3, and all of that — why would the Department produce tables for five grades, A* to D, as opposed to four grades, A* to C. That might send out a pretty powerful message to schools, pupils and employers about what is acceptable.
The point is that a grade D equates to a level 1 qualification. The Department for Employment and Learning, DETI, employers and others have told us that they value a level 2 qualification. So, we would be mixing the grades. Of course, we could do that and have much better stats.
Is the issue not that we are talking about one point at the boundary?
It is only one point at the boundary, but, for the vast majority of children, it is more than one point.
The gap between a top C grade and a D grade could be 18 points.
Mr B McCrea:
How wide is the band?
I will make three points. First, Basil, it could also be argued that there is only one mark between grade D and grade E.
Secondly, an interesting thing happens in awarding. As I said, the boundary between grade C and grade D is a judgemental boundary. Let us say that that boundary is at 112. Examiners will look at scripts from about 118 downwards and those coming up to 112. The area of uncertainty will probably centre around three or four marks. Examiners will look at the scripts and will determine which are grade Cs and which are grade Ds. A boundary has to be set.
Thirdly, there is no sense that we are putting pupils with grade Ds into a box, closing the box and saying that that is that. Pupils can move into essential skills qualifications, so I do not see it as a dead end of achievement at 16.
Dr Chris Hughes (Department of Education):
A sudden death element was mentioned. There is a difference between the number of year 12 students and school leavers who achieve grades A* to C. The figure for school leavers is higher by 4% or 5%, because those who do not get an A* to C grade in the first instance repeat. For example, 54% of year 12 pupils who sat English and maths in 2008-09 got A* to C grades, but 58% of school leavers left with A* to C grades and three other good GCSEs. It is not sudden death. People re-do the exam, and we have evidence that several percent of them get it.
Due to the qualifications framework that Roger talked about, the D grade will sit within the level 1 qualification. That will be an achievement in itself that creates a platform that can be built on to have another go at level 2. It is certainly not a cut off.
There are also the essential skills qualifications, so there a number of other platforms and options.
Mr B McCrea:
I may have missed part of the dialogue. When the Committee was in Omagh, I recall a number of school principals responding to Trevor’s questioning by saying that a D grade signifies some success and an ability to go out into the workplace and do all of those things. If you are telling me that that is not correct because of the level 1 and level 2 qualifications, we may have to address that issue. This is a very important matter. I take on board the fact that pupils can re-sit exams.
In our earlier discussion about PISA, we talked about how people value schools and whether they think that education is a good thing. It goes to the heart of the issue about labelling people as failures, which, regardless of academic assessment, can happen if one continually produces tables of the numbers who achieve five A* to C grades at GCSE. On a five-year average, 13% of pupils in the Belfast Boys’ Model achieve five A* to C grades at GCSE. In other schools, the figure is 90% or whatever. That affects people’s whole attitude about whether a school is successful, whether the pupils in it are successful, and whether this is right.
We are in danger of missing the boat when trying to encourage people by saying that they are successful. A clear delineation between the C grade and the D grade may be one thing. However, given the subjective nature of the way in which you are doing it, and the relatively narrow band for the marks, you may have to consider something in a more joined-up sense.
We are linked into a big Ofqual project about reliability, because, for example, the person I mentioned earlier who got 112 one day could have got 114 on another.
Mr B McCrea:
The issue is one of reliability.
That is a key issue emphasised to us by those who receive our school leavers around a level 2. We set targets for GCSE success at any grade, and at the A*-C because we recognise that A*-C equates to a level 2.
Mr B McCrea:
That is what worries me. There is an adage in computing: rubbish in, rubbish out. Our whole education strategy is based on Every School a Good School, raising standards and doing all those sorts of things, yet one looks at the disparity of the figures. I wonder what the ranking of schools here would be if a league table listed them according to the five grades A* to D. I suspect that you would have a completely different impression of how you need to proceed.
There are two points about that. First, if we accept the level 2 point, which is what all the folk responsible for taking students into the workplace and our economic departments tell us, then why would we not want more of pupils to be achieving at that level? Why would we not want to get those at grade D up to grade C, so that they are leaving with the full depth of knowledge and skills that the workplace says is needed?
Secondly, and this is an interesting point from a school improvement perspective: when we look at schools, we quite often see pupils achieving level 2 in history, geography, science, home economics and PE. The real issue is that if they are capable of achieving level 2 in a subject such as history, which is broadly literacy based, there is a strong argument that the vast majority will also be capable of achieving that level in English. Schools will be looking at the performance of pupils in individual areas and whether they can get to level 2 because that will give them an enhanced level of skills, and it seems to us that it is right to want as many pupils as possible to have the potential to achieve at that level.
I go back to the A* when it comes to A levels, or the AS and A2 levels as they are now called. You said that this was in response to universities needing to be able to distinguish. Are local universities accepting A*s? Are they graded for UCAS points?
It is up to the university how UCAS points are used. The line that Queen’s gave us, and the University of Ulster would be similar, was that they were monitoring the A* grade and would use it in some circumstances. They are now considering it for medicine. I can see, to some degree, why they do not want to base everything simply on academic achievement.
It is interesting that Oxford has now said that about maths and science. You can see that they would want the very best for maths and science, but medicine is a bit different. Queen’s is still looking at the issue, and medicine is a big issue for them. They are using objective interviews quite a lot.
There is a similar problem with physiotherapy at the University of Ulster. When it was introduced in England, almost immediately the modelling showed that the independent schools would get the A*s. At the same time, there was a big push by the Labour Government to get more of the lower ability — I forget the figure; the A, B, C, D — into universities. Therefore, there was quite a reluctance to use it.
There was the other dimension of the stretching and challenge aspect that we wanted to add to differentiate within the grade A high achievers and reward that high achievement. The jury is still out on how A* will be used in local universities.
Obviously, they could not use it for intake this year if they have not said that they would do so.
Queen’s said that they were looking at it this year, among other factors, in some circumstances. However, they certainly did not put it up in headlines in relation to their appointments.
The only difficulty that we have is that if we raise the matter with the DEL Committee, because it goes into another area, time may be against us. However, I am quite happy to raise this with the DEL Committee. Has the matter been conveyed to the DEL Committee, Roger?
Do you mean the A* grade and the universities?
Yes; and the whole interaction between that and what the universities are saying because it is an important issue.
We would be in discussion with DEL officials, and we talk regularly about that with the two universities. We also keep a monitoring role in universities and are involved with how it is being dealt with by the CAO system in the South because there were issues there. We are in ongoing discussions about that.
Is GCSE grade C benchmarked internationally in relation to the level 1 or level 2 boundary?
GCSE could be called a country-specific qualification. There is no benchmarking as such. One role that Ofqual has in its changed role in England is to look at GCE, which they say is university entrance, and the standard of university. They are looking at the different systems in England, Wales and Northern Ireland where about 75% of pupils take three subjects, and 10% take more subjects. In the South of Ireland, however, it is six subjects, and there is a broader range of subjects in Scotland. So, one could argue that they are taking a broader range of subjects to less depth whereas we take a narrower range of subjects to a greater depth.
That will come out in the initial work. It was interesting sitting in the Public Gallery when the Committee were discussing PISA. I see PISA as almost a dipstick comparison whereas CGSE is a qualification in which one is trying to cover a very broad specification, which says: “There is your specification; there is the nature of the areas that we want you to assess in, and we are assessing it.” However, there should be some relationship.
There is also the fact that GCSEs now plot on to the qualifications framework at level 2, which itself reads across at European level.
Does it read across?
Mrs Karen McCullough (Department of Education):
There is ISCED, which is an abbreviation for something that I just can recall at the moment. However, they tried to do it at Eurostat level and say how things equate but there are lots of question marks over it because one is not comparing like with like. Scotland specifically said: “Is it apples and bananas?” which is why they went for PISA because it is the same.
OK. Katrina, Roger, Chris and Karen, thank you very much. That was very useful and will make a contribution to our inquiry. Thank you.