Dukes Plus
Beyond the classroom
Dukes Plus Dukes Plus is a provider of university application support, life-changing career and university experiences, day camps, tuition, and summer schools.
University Applications Our curated collection of consultancies represent the best of Oxbridge, US, and Medicine application support.
Experiences Dukes Education is home to the UK's leading careers experiences, summer school and activity day camp providers.

75 Years of A Levels: How Questions Have Evolved

The year is 1951. Nat King Cole is in the charts. The Day the Earth Stood Still is in cinemas. Winston Churchill, aged 76, returns to Downing Street, and The Archers plays in the background as 37,000 students across the country revise for a new set of exams. The Advanced Level was a new way of identifying pupils for university, taken by the minority who stayed at school beyond sixteen. In many ways, the basic purpose of the A-level is much the same now as it was in 1951, but in substance it has evolved a great deal since then. 

Let’s take a look at the sort of exam questions that lay in wait for the first candidates. Perhaps most striking to the modern reader is how broad the syllabus could be compared to today’s papers. The history curriculum provides the most obvious example; the 1957 paper (the first publicly available) roams across vast stretches of spacetime, down avenues that would certainly tax the modern student, perhaps even the modern history buff. “What effect had the Lombard conquests in Italy on the history of the Papacy?” asks Paper IV (European History, 800-1494). “Account for the contrast between Prussian and Polish history during the seventeenth century,” asks another. 

The French paper was similarly essay focused. Alongside translating texts, students were expected to “Discuss La Fontaine’s portrayal of a few of the social types of his time” (in English), “Write a critical appreciation of Vigny as a lyric poet” or even “Describe the terms of the Locarno Pact (1925) and estimate its importance to France”, rather than crane their ears, as today’s students must, to ascertain where Jean-Pierre went on his holiday. Listening (or, for that matter, spoken) exams did not feature at all until the 1970s.  

Students looking to escape the dreaded essay found no relief in the sciences. The 1957 biology paper gave students a choice of five meaty, wonderfully general questions. “Write an essay on sugars in plants”, “Discuss the cell concept in plants and animals”, or explain – in a question glinting with philosophical temptation – “In what ways does social life in insects differ from the social life of man?” 

Those more squeamish candidates may take little joy in the Practical Test (three hours), which asks them to “Dissect the frog provided, K1, to display (a) the alimentary canal, (b) the glands associated with the alimentary canal, (c) the hepatic portal system”. Candidates must then excavate “the median portion of the pectoral girdle”, followed by “the lobes of the liver, the stomach and the small intestine”. Once the frog has been entirely disassembled, students are asked to sketch the carcass (cross-curricular education at play) and label its parts.  

To the small relief of frog lovers, A-levels were still in their infancy in 1957, with only 4,000 students taking the biology exam. 75 years ago, only 3% of students took A-levels (this summer it was 43%). They were meant for an elite minority heading for one of only 17 UK universities. 

The first A-levels were pass or fail only. Swots could expect a ‘distinction’ grade from 1953 onwards, but the familiar A-E didn’t appear until the 1960s. Originally, grades were proportional (the top 10% got an A, the next 15% a B, and so on), but by the 1980s the idea that they should reflect a fixed standard achievement took over, ushering in the ‘grade inflation’ and ‘dumbing down’ debates so familiar in broadsheet columns today.  

The original qualification put a heavy emphasis on recollection and regurgitation. Even STEM students might be expected to produce several yards of writing in a given afternoon. Whilst essays still matter today, A-levels are now far more skills-based. The vagueness has evaporated, replaced by more rigid questions with clear explanations of what is expected from the student. ‘Write an essay on sugars in plants’ becomes ‘Sugars are important in plants. Explain (a) the role of sugars in plants and (b) how they are transported around the plant, with reference to photosynthesis [6 marks]’. Questions are tidier, with more sub-clauses and prompts for specified evidence, such scaffolding perhaps anticipating, in infancy, the walls that will be placed around the candidate’s increasingly specialised job.  

And today’s most popular subject? Maths, with over 115,000 students sitting the exam this summer. The main beneficiary of the recent STEM boom, maths is followed by psychology (79,000), biology (74,000) and chemistry (66,000). Maths exams, perhaps unsurprisingly, have changed little since the 1950s, bar the absence of the trusty calculator, a 1970s innovation. Psychology, economics and computing have exploded from relative obscurity in the last decade, whilst languages continue to decline (only 6,500 students took French this year; 2,200 took German). 

But as we look at the future of A-levels, we come to the inevitable discussion of AI, which made itself at home in many students’ lives. We all know that ChatGPT would make short work of a 1950s question, just as it does today’s assessments. Ofqual, the exams regulator, pounced on roughly 3,000 cases of recorded malpractice last year deducting marks or even removing qualifications from those who failed to cover their tracks convincingly.  

But is proscription enough? Does the A-level, in its 75th year, look a little feeble in the face of this nimble technological hydra? Perhaps not. The examination hall, with its phoneless isolation, seems to have a healthy future. Coursework that can be written (or generated) remotely may recede into obscurity, but there is inherent value to knowledge and its unassisted recollection, or to the forming of a coherent argument. The ‘what,’ ‘when’ and ‘why’ of the French Revolution aren’t just nice to know, but helpful in making some sense of the perplexing world we find ourselves in. There is a reason we choose to cycle without stabilisers. 

Perhaps there is a sense in which AI makes knowledge even more valuable. How can someone meaningfully interrogate ChatGPT about the French Revolution if they know nothing about it? It may be a case of how questions are phrased. For example: 

“Here are five competing explanations for the French Revolution. Which is the most convincing, and why?” 

“AI has produced the following explanation for the recent unrest in Serbia. Identify and explain its three most significant weaknesses.” 

But what about the sciences? Should the biologists return to their frogs? Perhaps there is more value in experimental judgement, giving students an opportunity to behave like actual scientists. 

 “What is the most serious limitation of this experiment?”  

“An AI has interpreted these results as demonstrating X. Is this conclusion justified?” 

A-levels were designed on the premise that knowing vast amounts of information matters. It still does, of course; the question for the next 75 years is how we can teach students to navigate an endless sea of it, to distinguish from the noise, to fish evidence out of assertions, and start questioning answers as well as they can answer questions.