Are we assessing the right things in the right way in MFL? Sadie Thompson argues that we aren’t. There is a better way…

The current GCSE assessments for Modern Foreign Languages in England are facing upheaval as the specifications for French, German and Spanish are set to be replaced for first teaching in 2024, with the first of the new exams due to be taken in 2026.

Although these changes are in their infancy and yet to be set in stone in terms of an assessment model, it provides an important point to reflect on assessment in MFL currently and how to design an accurate model which rewards authentic communication and ensures accurate grades for all learners.

What are we striving to achieve?

As a linguist, my primary aim is to ignite a lifelong love for language learning amongst my pupils and to create linguists. Idealist though this may be, secondary to that comes the need for my students to pass exams and secure highest possible outcomes; Im usually banking on that being a natural and welcome outcome.

Yet all too often this is not necessarily the case and talented linguists, indeed sometimes even native speakers who enter for an extra bonus exam grade, can surprisingly lose marks. This leads us to question our assessment model and what we are striving to achieve in the first place. Do we want great results or linguists for life?

Are we expecting too much?

As most MFL teachers across the country will tell you, the current GCSE specification falls short. There is too much content to teach across a two-year key stage 4 and too many exam-specific skills to master.

The assessment is based on 4 key areas, Listening, Speaking, Reading, and Writing, each comprising 25% of the mark and with no allowance given for the mixing of tiers to play to studentsstrengths. It is therefore vital to address the skill of answering exam questions successfully.

Unfortunately, this is too often at the expense of valuable curriculum time which needs to be devoted to an already crammed-full-with-content specification. Losing the ability to mix tiers is also problematic in that many MFL classrooms are mixed ability. It is also unusual for a given class to be either entirely higher or foundation candidates. So the teacher must also become an expert in teaching a content-heavy specification, while addressing exam skills for groups of students at two different tiers of entry.

Add in the fact that the two tiers then have vastly different mark schemes for Speaking and Writing and you are creating an almost impossible challenge for any classroom teacher to rise to.

Natural speaker or exam performer?

It is no secret that most GCSE candidates dislike the Speaking exam with a passion. In fact, most adults can recollect most details of the toe-curling experience, some can even recite a line or two twenty years later! However, this is probably because the scenario in itself is so high-pressured, so daunting and presents so many hoops to jump through that it never resembles a real-life conversation.

Even the so-called General Conversationcomes with a list of structures to try and shoehorn in for top marks in the Range and Accuracy of Language category, all the while sounding natural enough to get full marks in the Spontaneity and Fluency category. Confused? Thats before weve even discussed the other 30 marks available for the Role Play and Photo Card elements of the test.

As a linguist first and foremost, this has never sat right with me. My ambition in teaching is to inspire a generation of future linguists. We want pupils who go out into the world with a confidence and an ability to hold a conversation in the target language, at least enough to leave them wanting more and motivated to continue learning so that their next conversation is always better than the last.

I suspect this in the same way a History teacher might want to inspire future historians, or a Food Technology teacher wants to instil a lifetime appreciation for delicious and nutritious food. However, the real frustration is that the Speaking assessment for the current GCSE is not suited to a good natural linguist, but is suited more to someone who is good at memorising chunks of language, and therefore “performing” in exams.

I have known native speakers to drop marks because they did not include enough examples of the future tense, favouring the natural conjecture of the conditional, no matter how accurate or fluent their response. This isnt right. In the assessment, a native speaker should be able to achieve full marks with minimal effort.

Recently on Edu-Twitter…

In a recent #MFLChat (7/2/2022), the weekly online Twitter chat for MFL teachers, we were asked to imagine if there were no more exams, what lessons would look like. The responses centred mainly on two key themes: incorporating more culture in lessons through the inclusion of music or literature, and a heavier focus on speaking and communication.

We can deduce then that it is not that teachers want to spend less time concentrating on speaking but more that they want to be able to devote more time to more authentic communication. At the very least, teachers want their students to be rewarded for authentic communication in the exam without the need to artificially squeeze in grammar features that don’t occur in real-life conversations.

This is maybe a fault with the mark schemes. The role-play element of the exam is arguably redundant as time goes on, presenting exchanges such as booking a hotel room or buying concert tickets which would more than likely be done online without the need for any interaction in the target language. But the photo-card elements of description and sharing opinions and a general conversation across several themes seem sensible and appropriate to enable candidates to develop answers and share a range of structures. It is the demands of the mark scheme which requires the jumping through hoops.

What do we need?

We need an assessment model which is in line with age-related expectations, given curriculum constraints, and that also rewards the kinds of real-life interactions that would be useful in real life.

This is probably a conundrum faced by exam boards for all subject areas – how can we authentically test a wide spectrum of knowledge, whilst showing progression over time, but also instil the key skills and attributes of those expert learners? Is it ever possible to create an assessment which rewards those skills of a true historian, a real mathematician?

Im mindful that the perfect assessment model for languages probably doesnt exist and that exam boards, restricted by requirements set out by Ofqual, can only seek to change so much when designing exam content and marking criteria.

However, what is painfully apparent is that GCSE results in MFL are historically far lower than other options subjects. How can this be, when it is the same students taking the exams? Surely, we arent producing a nation of students that can access the material of a Geography paper and secure a grade 9 but cant replicate that success in French?

We must ask questions of the current assessment processes, to ensure that in the first instance, our students arent short-changed out of the grades theyve worked hard for, nor that they are discouraged from the rich and diverse experiences learning a language offers in place of securing better grades elsewhere.

We need to level the playing field and that doesnt have to be through starting again but tweaking what we do have and ensuring it serves our purposes best.


Sadie Thompson is Head of German at Thornden School in Hampshire. She has been teaching for 11 years and as well as flying the flag for German, she also teaches Spanish to Key Stages 3 and 4. Sadie has a keen interest in educational research and is currently Deputy Director of HISP Research School.

Write A Comment