top of page

Estimating and Ranking in FE

On 18th March, the government announced that all GCSE and A level exams would be cancelled for 2020. In the frenzied days afterwards we all wondered what the next steps would be and how we would (or could) follow them.

Well, in the time in between, we’ve had lots of answers, followed by more questions, followed by some answers.

And here’s what we know:

- Teachers should use all available evidence to provide an estimated grade which they feel students would likely have achieved;

- Teachers will also rank order students within grades which will be used to rank order all students within each subject/ qualification, within each centre;

- This will be submitted to the DfE starting from 1st June and grades will be calculated using this data and a range of other data/ information.

This process isn’t easy for anyone, but I would say it is simpler for some schools. Y11 cohorts will likely have had two or more teachers in Y10 and Y11, a range of the required evidence and, for cohorts of 150-200 students, a smaller number of rankings to complete.

For an efficient and concise example of how to complete the process, I would direct you to this excellent blog:

For FE practitioners, read on.

In Further Education, some resit cohorts number more than 5000 students. How do you rank 5000 students? This is without taking into account the challenges of English and maths in FE – lessons of 3 hours per week for GCSEs, E&M attendance of around 75-80% and the difficulty of gathering relevant evidence (some colleges can’t hold ‘traditional’ mocks due to massive student numbers).

Firstly, what are the responsibilities within the process?

For Department Heads:

- Analyse fairness and accuracy of the allocated teacher grades through detailed consideration of student starting points and predictions, progress review data, key assessment data and the mock examination data;

- Where issues arise with this grade estimation/ ranking process, the E&M HoD will discuss and collaborate with staff to amend any issues (this includes any students who are given the same estimated grade and mark);

- Provide all ranked cohort data in line with government timelines and the above proposal;

- Responsibility for overseeing and maintaining the efficiency of the above processes;

For Practitioners

- Provide a timely and (as best as possible) accurate grade estimation and ranking for all relevant students using all relevant data and evidence (as outlined by Ofqual and the DfE);

- Engage fully with all relevant tasks as outlined in this document and within E&M;

- Engage in a full series of feedback and collaborative discussions with relevant stakeholders (where appropriate) including students, parents/ carers, HoD and senior leaders (ensuring only stakeholder relevant information is communicated).

But how do we start the process? An accessible starting point, is to break the process down into the following steps:

1. Estimate grades;

2. Rank within that grade;

3. Collate an initial whole-centre draft;

4. Calculate centre and teacher estimated ‘results’ and compare to previous years;

5. Focus on ‘clumps’ and hold meetings to discuss further ranking.

So, let’s look at each stage…

1. Estimate grades

The simplest stage because it’s completed constantly, throughout the year and requested at this time anyway. As DfE and Ofqual guidance outlines, this is done using all available evidence – mocks, work done in class, homework, initial assessment data etc. In FE, it has even discussed that this could potentially even stretch to attendance and how this would have impacted on a final grade (in-class attendance and attendance at revision sessions).

2. Ranking

Each teacher should also rank their students within each grade. How do we do this? Some centres have used decimalisation for this process. For example, if a student is estimated a grade 5, a rating from 1 – 9 will be added after the grade to represent how secure that student is within the grade. A student who is extremely secure at grade 5 will be given a 5.9, with the least secure awarded a 5.1. Where there are several students given the same ‘rating’, a second decimal is added (a student with 5.95 would be above a student with a 5.9).

A great approach, but is not without issue. Standardising and moderating without lengthy meetings can be difficult and even though exemplar student work can be produced, this can take time and there is still room for error among more confident or inexperienced teachers.

So, what else can we do? Why not tap into and use existing systems, which are already used to rank and grade students? Raw exam marks. For GCSE English, students are awarded marks out of 160 and for GCSE maths, 240. Why not use these numbers to rank students?

The exemplar material already exists and is constantly distributed, with standardisation and moderation occurring throughout the year as part of normal CPD (with the majority of staff having worked as or having close contact with an examiner). The available evidence is easily transferrable to this system with a potential template below:

· For Functional Skills, this layout would slightly change:

- For maths it would remain broadly similar:

- However, for English, there are several components to be incorporated:

By no means a flawless system, with issues being obvious, it does facilitate an efficient whole-centre draft. Collate all teacher data onto a spreadsheet, sort raw marks from highest to lowest and this is an instant first draft (as long as everyone is using the same grade boundaries – easiest to go with the 2019 boundary).

A friend commented that this was extremely granular, and asked if staff would be able to rate/ rank at this level. I agree, but the new system demands that we be this granular and this proposed system uses information and subject knowledge already deeply embedded with practitioners.

3. Collate an initial whole-centre draft;

This next step is overseen by the HoD (or leader) for E&M. Keep templates as simple as possible at this stage so information can easily be transferred to the DfE/ exam board template when it comes. As the above example shows, it is best to use only the most essential information, including learner name, college number, estimated grade, mark and rank. As above, to get a ‘first draft’ ranking, sort the list of students by raw mark (highest to lowest) and you have a starting point.

4. Calculate centre and teacher estimated ‘results’ and compare to previous years;

This is vital as we know previous centre achievement will be used. It is not to say centres cannot achieve great leaps in year on year achievement, it’s just that it is unusual.

Having said that, this is possible in FE as each new cohort truly is different to the last. Also, if a centre can evidence a large jump in results, there should be no issue (just make sure the evidence is plentiful and easily available).

First, see if your whole-college predictions fit within the centre trend based on previous results. If you’ve shown a modest increase from 2018 to 2019, it is reasonable to expect this for 2020. If you’ve seen a modest decrease from 2018 to 2019 and (for example) are expecting to double achievement for grade 4+ in one area, investigate further.

Look at individual staff predictions - it could be that one teacher’s enthusiastic estimates are impacting the entire prediction. This is also a great tool to combat teacher bias – more on that here from the wonderful people at Pupil Progress.

Again, this is not to say that anything must change (as long as estimates can be evidenced, no change is required), but it is important to ask the question and be thorough. Double check – make sure there is evidence.

5. Focus on ‘clumps’ and hold meetings to discuss further ranking

The Hard Part. It is likely there will be several students with the same raw mark. How do you split the students on 83 marks? Firstly, organise meetings with each relevant team (and only relevant staff – don’t make the maths team attend the English portion of the meeting).

Prior to meeting, send teachers lists of grouped students and ask them to prepare, identify and collect evidence which supports the mark and estimate they have given and bring this to the (probably virtual) meeting. It’s best that this work is viewed by everyone and the HoD makes the final decision.

The responsibility for making the final decision should always be with the HoD.

It is important to have these meetings and have transparency throughout this stage, because there is will be a cut off for each grade. On results day, it will be as straight forward as ‘Student 45’ achieving a grade 4 and ‘Student 46’ achieving a 3. This is why you must make sure that student 45 and 46 are ranked correctly.

And finally...

Again, none of the above is a replacement for the process described in the blog link at the start. That process is highly effective and there is no better way than meeting up and looking at the evidence. However, this takes time and resources which might not be available and can be impossible for leaders with thousands of students.

For FE specifically, this is a difficult process simply because we encounter many more challenges in prediction: the students with poor attendance who achieve a grade 5, the amazing students who put in every effort, every day and continue to achieve a grade 3 (and everyone in between).

Again, this outline is by no means perfect, but it is a system and it gives students an opportunity to achieve based on their past performance and potential – something we should all support.

I’ll be sharing this on Twitter soon, and would ask anyone who finds fault to share constructive criticism kindly and take the above in the spirit it was intended.


Recent Posts

See All


bottom of page