Why schools still can’t (easily) measure skills

As part of a wide-ranging review looking into why so many young people are struggling to get into employment, former Labour cabinet minister Alan Milburn revealed in his interim findings that nearly three in four teachers believe the current curriculum puts too much emphasis on passing exams - and not enough on some of the skills needed for work.

With almost 1 million young people aged 16 to 24 in the UK currently not in education, employment or training (Neet), there is hope among many that a renewed emphasis on skills in revised curriculum programmes of study, prompted by education secretary Bridget Phillipson, could help to fix that.

However, the idea that the curriculum could become more skills-heavy has left others concerned - not just because defining skills is tricky, but because measuring how well they are taught is fraught with difficulty. That brings associated dangers for heads in terms of accountability.

So, how far are we able to properly define skills, and how valid are those fears over measurement?

What are skills?

There is not as much contention around the definition of a skill as you might expect, considering the knowledge versus skills war has raged for decades (see box below).

Bill Lucas and Guy Claxton, professors at the University of Winchester, have defined a skill as a “learned capacity to do something useful”.

They explain that there are skills like reading and writing, but also oracy, creative and critical thinking, problem solving, and many more.

Lucas argues there are at least four types of skills: cognitive; metacognitive; social and emotional; and physical and practical.

According to Christian Bokhove, professor of maths education at the University of Southampton, another way to think about “skills” is as practical or procedural knowledge - ie, how you do things.

This is different to “declarative knowledge”, which is what we more traditionally think of as “knowledge” in the curriculum - for example, the year a certain historical event happened.

Daisy Christodoulou, director of education at No More Marking, admits that the distinction between skills and knowledge is not clear-cut. She says that to teach complex skills, we must break them down into smaller parts - which, in many subjects, will look like knowledge.

“Opposing knowledge and skills is like opposing ingredients and cake - the ingredients make the cake, the knowledge makes the skill,” she says.

Lucas similarly believes that knowledge and skills are interconnected: “Knowledge-rich and skills-rich are simply two sides of the same coin; all knowledge is made up of skills of various kinds and complexities.”

The challenge of measuring skills

Given this overlap, how well can skills be assessed? It depends on the skill, most experts say.

For example, the current curriculum already measures skills like reading and writing well, says Christodoulou.

Where it gets more difficult, she argues, is when we try to think about skills that cut across multiple different domains.

She says that for creative thinking, for example, if we want students to be creative in many different domains, they would need “quite extensive content knowledge in all those different domains” - and this then becomes difficult to assess.

Lucas also says that how well we can measure a skill depends on how expansive that skill is.

“If I took the example of reading a map. That’s a relatively definable skill, which I could come up with a number of practical tests for,” he says.

There’s also relatively good evidence that we can measure speaking and listening, as recent thinking about oracy has shown, he adds.

But “if you look globally at the kind of skills that educators want to assess, they are quite often described as the four Cs - creativity, critical thinking, collaboration and communication - and all of those are quite complex,” he adds, and “the more complex the skill is, the more tricky it is to assess it and the more it needs to be broken down into constituent elements”.

So, what would a test that measures those more expansive skills look like?

Models for measuring skills

David Thomas, a former government adviser who is currently CEO at Axiom Maths, suggests that when people talk about skills, what they often mean is the application of knowledge in environments that are less controlled and less rehearsed than the exam hall.

To get a measure of a child’s skills, it would therefore be helpful to look at whether they can apply their knowledge in complex settings.

The problem, he says, is that generating complex or unpredictable settings for assessing skills across domains is very difficult.

He gives the example of trying to assess whether someone is good at teamwork, using a rubric.

“You write a rubric that says a grade one team worker understands peers poorly, and a grade two understands them acceptably, and a grade three one understands them well, and a grade four one understands them excellently, and then you just expect people to award grades based on that set of adjectives,” he explains.

These criteria, he continues, “don’t mean the same to every person, so it’s very hard to in any way construct a reliable measure”.

An alternative, he suggests, might be breaking a skill down into a list of very specific behaviours. But this is not perfect, either, as what tends to happen is that “teachers teach children these very specific behaviours, and then what you get is children performing something that apparently is teamwork that looks nothing like what anyone [outside the classroom] would possibly recognise as teamwork”.

Lucas has had experience of trying to put some of this into practice; he co-chaired the development of the Programme for International Student Assessment (Pisa) 2022 assessment of creative thinking.

Those sitting the test were presented with a number of scenarios that asked them, for example, to come up with ideas to improve an initiative that promotes carpooling, or to come up with a story idea when presented with the cover of a book.

A test like this allows students to demonstrate their level of creativity, but there is still the challenge of deciding what a good answer looks like and training markers to recognise it - a process that is “inevitably cumbersome and expensive”, Lucas says.

While finding ways to assess skills might not be easy, it is important, he adds, because it increases the skills’ status.

You can put skills on the curriculum without measuring them, but research suggests that the quality of teaching and learning will be higher with clear definitions and suggestions of progressions and an expectation that skills will be evidenced.

The stakes of any test of skills would be important, though, Bokhove points out. He says assessments would need to become far more reliable before they could be used to hold schools to account.

And for teachers to be able to measure skills reliably, there are other things that would need to change, too.

A skills-based curriculum?

Thomas suggests that opportunities for children to explore applying subject skills in more unpredictable situations currently get crowded out because examined content dominates the curriculum - something that the Milburn review also points to.

The government did engage with this idea in its response to the curriculum and assessment review, committing to reducing GCSE exam time by 2.5 to 3 hours on average.

However, many specialist provisions are already delivering skills-based curricula and measuring their students’ progress in those skills as part of this.

At Nigel Bowes Academy, a special school for primary-aged children with social, emotional and mental health difficulties, lessons are structured around modelling and practice of skills, with progress measured through a mixture of formative assessment, practical demonstration, observation and questioning, says assistant vice principal Caroline Allen.

In another model, at special school Longspee Secondary Academy, teachers use the “gradual release of responsibility” approach (a structured teaching model centred on slowly devolving responsibility from the teacher to the learner) to model and embed skills.

Progress is then measured through teacher observation (often with the support of trained teaching assistants) and end-of-unit assessments, in which students are asked, for example, to create a visual representation of learning or teach the teacher what they have learned.

Some special schools use a framework developed by Skills Builder, which breaks down eight skills (problem solving, listening and speaking, for example) into steps for teaching and uses benchmarks to measure progress.

Meanwhile, in mainstream schools, there is another method already being used, which could potentially be scaled up: the extended project qualification (EPQ).

The EPQ is an independent project in which students plan and carry out their own research on a topic. It can result in a written report or something more practical like an artefact or production.

“I would make that compulsory for everyone,” says Peter Hyman, former headteacher of School 21 and adviser to Tony Blair.

“It teaches you initiative, research skills, critical thinking and project management. You’re doing all those skills in one project like problem solving, overcoming obstacles and resilience.

“That is one very simple thing that already exists as a qualification that is done at the margins that could be moved to the mainstream.”

Hyman also says many employers already do strength-based assessments on various criteria that could be used in schools - but further assessment reform would be needed for this to work, as well as alignment of incentives.

“Can schools do it? Yes,” says Lucas. But not if they have too much to do.

As a starting point, he suggests taking “one or two of the areas where there’s a good proof of concept, like creative thinking” and considering how those existing tests can be appropriately adapted and embedded in a number of subjects, before prototyping the model.

He adds that he would not “inflict” this on the teaching profession without the appropriate resources, though, and goes on to reiterate Bokhove’s point about the stakes attached to a new assessment of skills.

“It doesn’t have to be perfect, only good enough - as long as it isn’t high stakes,” he concludes.

Tes asked the DfE whether skills delivery would be reflected in accountability frameworks in the future. The DfE did not respond.