Problems with Replacing GCSE Computer Science

It's been an interesting few days since the BCS video call presenting three ideas for replacing GCSE Computer Science.  In trying to form a useful opinion, I have had very rewarding discussions with colleagues from many subjects and reflected on these and my own experiences.  I am extremely grateful for the prompt to seek out those opportunities - thank you to everyone that participated.

Even after all my efforts, the only thing I am completely certain of is that this is a phenomenally complex issue.  However, I would like to share some of my thoughts on those complexities in the hope that we can avoid at least some of the previous pitfalls.

Before I continue, I should point out that any opinions below are my own, not those of any employer or organisation.  I have tried to express my thoughts plainly whilst avoiding being blunt or glib.  I'm not sure to what extent I have succeeded.

The Current GCSE

In my personal opinion...

• BAD: it does not prepare students well for A-Level Computer Science because there is too much recall and not enough practical problem solving.
• NOT NECESSARILY BAD: it does not include creative tech projects because there are other Level 2 courses which are meant to cover that (e.g. Cambridge Technical IT, Creative iMedia)
• REALLY BAD: it in no way prepares students for the collaborative nature of working as a programmer.

Themes from the suggested replacements

Before trying to replace GCSE Computer Science, we really need to look at the whole range of qualifications currently on offer.  If you think the other qualifications are failing, then trying to turn the GCSE into one of them will also fail.

Teaching Computer Science Through Context / Creative Projects

Cambridge National IT (https://www.ocr.org.uk/qualifications/cambridge-technicals/information-technology/qualifications-at-a-glance/) already has units on cyber security, web development, games development, data analysis, networking, etc., etc.  I have taught the Level 3 qualification before, and it gives lots of great opportunities for practical development and problem-solving.
So whilst I have not taught the Level 2 myself, I believe the question still remains: why make GCSE Computer Science more practical when the Cambridge Technical is right there already?

Creative iMedia (https://www.ocr.org.uk/qualifications/cambridge-nationals/creative-imedia-level-1-2-j834/specification-at-a-glance/) already covers a range of practical creative projects.
I have taught the Level 2 qualification before.  Whilst it is not perfect, it is a great course.  Why make GCSE Computer Science involve creative practical projects when Creative iMedia is right there already?

If your answer to both of those qualifications is that they are 'vocational' or 'not academic', then please remember there are university degrees in 'Computer Science' - it is an academic subject.  I would ask you to consider the question: would you suggest we replace the GCSE sciences with creative engineering qualifications?

If you are thinking along the lines of 'we need people who can actually program in the workplace' then perhaps we should try to fix the issues with the Cambridge Nationals instead?
(although I don't think the problems with Cambridge Nationals are actually anything to do with the courses themselves)

On the other hand, if you think Computer Science is actually a practical subject not an academic one, then I would ask you to consider why we have a GCSE for it at all.

I think practical skills and creativity are absolutely vital in programming.  I also think 'Computer Science' is not the same as 'programming'.

To my mind, the big elephant in the room here is: why do people think these other Level 2 qualifications are not as good as GCSEs?

Teaching Collaborative Skills in Computer Science

In my opinion, this is the biggest problem with the U.K. education system as a whole.  Pretty much every workplace is a collaborative environment, and I see nothing that even tries to prepare students for this.  It is truly the result of the assessment tail wagging the education dog.

I have been told by a colleague that part of GCSE Music involves a group performance.  Each student gets a mark for their part of the performance.  It can even be recorded multiple times and the mark for each student can come from a different recording of the same piece.  It surprised me to realise there is already precedent in a GCSE for individual assessment in group situations.

Even academia is collaborative - surely there is some scope for collaborative projects to contribute to students' assessment in GCSE Computer Science?

That, however, raises many questions about the administration and marking of centre-assessed work.

Some Issues with Centre-Assessed Work

The issues are legion, I have only considered a few.

In GCSE Art, students produce a portfolio of work across the two years.  It does not all have to be completed under high-controlled conditions, but the teacher has to see enough of the work being done to be confident the student has actually produced all of it.  So having students collect a portfolio of work they have done, even from collaborative projects, could provide mark-worthy evidence.

The grade boundaries for centre-assessed units end up being incredibly high because of the pressures on teachers to 'produce' good grades.

The underlying evil of school league tables makes school managers pressure teachers into giving high marks.  In a previous school, I was denied UPS pay progression because the GCSE grades were not good enough.  Illegal?  I think so.  Worth fighting?  I didn't feel up to it - I just left the school.

The competition for university and apprenticeship places makes parents and students pressure teachers into giving high marks.  I know of teachers who face formal appeals purely because the student hasn't worked hard enough and wants to make it the teacher's fault.

If the professional judgement of teachers is undermined instead of respected and defended, getting accurate marks on centre-assessed work is surely a lottery of management styles and student/parent demographics.

Even if teachers are confident enough to mark accurately despite the external pressures, mark schemes for centre-assessed work can be infuriatingly vague.  The AQA A-Level NEA has two beautiful examples in the 'completeness' category: 'A system which meets almost all of the requirements...' and 'A system which meets many of the requirements but not all'.  The difference between those two sentences is 5 marks out of 75.  In 2024, 6 marks could have turned a B into an A* on the NEA component.

When I taught Creative iMedia, students had to research and analyse a 'wide range of sources'.  In one unit, it turned out a 'wide range' meant 5.  In a different unit, it meant 3.  It can take years of back-and-forth between teachers and moderators before anyone is confident of exactly what these mark schemes actually mean.

Consider two notional assessment criteria:

• 'described, compared and evaluated a wide range of sources' 
• 'described at least 3 key features of a source; compared at least 2 key features of 2 different sources; evaluated at least 3 key features of any source against the needs of the target audience'

Whilst the second would still need clarification on what 'key features' are and what counts as valid 'needs of the target audience' (which could be provided in a simple list), the second is going to be so much easier to standardise across centres.

If you think providing numbers will undermine your method of assessment, please ask yourself: do you even know what you are trying to assess?  The reality is that defacto numbers will end up being defined anyway - it will just take a few years of students' grades getting trashed by moderators to get there.

Another issue with centre-assessed work is that teachers are forbidden from 'helping' (a.k.a. 'teaching') the students while they are producing that work.  The more time spent on the assessed work, the less time teachers have to actually teach.  I make Year 12 A-Level students do practical projects along the lines of mini-NEAs, writing objectives for one, doing formal testing for another, etc.  This is how I prepare them for the onslaught of the NEA where I am extremely restricted in the support I can provide.  It takes a huge amount of time investment to do that for what is arguably the most important unit but just 20% of the marks.

A Lesson from A-Level Sciences?

A-Level science qualifications have 'practical endorsements' (e.g. https://filestore.aqa.org.uk/resources/physics/AQA-7407-7408-PRACTICAL-ENDORSEMENT.PDF).  These to not affect the overall grade, but enforce the teaching and assessment of specific skills which are deemed essential to the subject.  Students also get multiple chances to demonstrate they are proficient at each competency.

Something along these lines in GCSE Computer Science could provide a framework to support the teaching and assessment of practical skills whilst also freeing teachers from the numerous pressures undermining their professional judgement.  

Students could build a portfolio through a series of collaborative group projects.  If a student is falling short, the teacher can step in to help, knowing that the student will have another opportunity to demonstrate the competency in a subsequent project.  

The exam papers can still differentiate the students on their academic ability, but the practical endorsement can go some way to show they can actually produce practical solutions to problems.

In Conclusion

It is clear to me that producing the suggestions for the BCS call involved a huge amount of effort and thought from many people.  They are honest attempts to improve the current situation and I am very grateful for the prompt to spend time properly thinking on these issues.

I fear that in trying to 'fix' GCSE Computer Science, we will end up duplicating previous mistakes.

Hopefully what I have written is plain and direct enough to be thought-provoking without causing offence or insult to those who have worked so hard already.

Update 19th May 2025:

Trying to make sense of my own confused / hypocritical ideas

I am having trouble answering this question:

If I think Computer Science is such an academic subject, why do I think the NEA is such an important part of the A-Level?

I value the NEA because it gives students a chance to build a real piece of software to solve a real problem.  I think that the experience they gain going through that process serves as an excellent foundation from which to further explore developing creative and robust software solutions.

That is clearly a practical engineering process and does not map well to the content of some 'Computer Science' degrees.  Should this practical aspect be in a different 'non-academic' qualification?

My understanding of GCSEs being 'academic' qualifications is challenged by at least two GCSEs which show that 'academic' qualifications can include 'practical' skills.

Design and Technology GCSE includes a portfolio which involves designing a solution to a problem, gathering user requirements and developing ideas into a final design.  A separate unit also involves some practical work with materials and tools.

AQA Engineering GCSE also has an NEA component involving problem analysis, problem solving, modelling and production of a prototype.  Unfortunately I have never met anyone who has taught this at GCSE so I have no idea how effective it is.

There is also a Cambridge Technical Level 2/3 in Engineering, which makes me even more confused about where the boundary is supposed to be.

A GCSE Computer Science with a software engineering component actually sounds pretty good to me (until I imagine what an NEA markscheme might look like for it).