Two eight-year-olds are quietly discussing the development of their “learning power”. “Six months ago,” says Sophie, “I was quite a good Team Learner, but my Stickability was pretty weak, I didn’t use my Imagination much, except when I was told to, and I never Planned what resources I might need before I started in on a piece of learning.

“Now my Planning and Imagining are much stronger - but I still tend to drift off when I don’t immediately know how to tackle something. I think I’ll make Stickability my LP (learning power) target for the next half-term.”

Rashida nods. “Yeah, I still get fazed when I’m confused, too, but I think I’m going to build up my Questioning next.”

Sound implausible? Not to these girls and their classmates. They have become so used to thinking and talking about themselves as developing learners that such conversations have become second nature. Their teacher is one of a growing band of “learning coaches”, who coax students to think about how to develop their “learning muscles” and “learning stamina”.

Assessment for the 21st century is changing its emphasis. Sure, we need to know how well children understand and retain the basics. But we also need tools to help teachers and pupils focus on and track their development as effective learners - not just in school, but throughout their lives.

Several prototypes - mostly self-report questionnaires - are already in use.

Rashida and Sophie, for example, are prompted to take stock of their own “learnability” by filling in a short questionnaire that contains statements such as “I know what to do when I get stuck with my work”, “I’m good at learning with other people”, or “I can really concentrate when I need to”.

They think about how well each statement applies to them, and rate themselves on a four-point scale from “rarely or never” through “sometimes” to “mostly or always”.

Then they compare their responses with what they said last time. In this way they can see how their different learning muscles are coming on, and can plan the next stage of their “training” in conjunction with their coach, the teacher. It’s motivating, it’s interesting, and it seems to work.

Being an effective learner requires more than the right skills. It involves attitudes, interests and emotional tolerances - for example, coping with confusion - as much as knowing how to use a “mind map” to organise knowledge, or mnemonics to remember French verbs. It’s no use being an effective learner if you don’t enjoy learning.

So we need to develop ways of tracking the development of these learning dispositions. Using questionnaires does have flaws: do students really see themselves clearly? Are they giving the answers they think their teachers want? But they can still be useful, especially if the students are actively involved in analysing and constructing the instruments for themselves.

A problem with traditional “personality inventories” is that they assume personality traits to be much more stable than they really are. Do I carry on in the face of difficulties? Well, usually in reading philosophy, but rarely when there are glitches in my computer. I’m more likely to persist with the gentle encouragement of my partner, and less after a shattering day. Does Jake ask questions? Well, loads in design and technology and hardly ever in maths.

As students develop and consolidate learning dispositions, they become less fragile and variable and develop more robust and wide-ranging ways of meeting the world. So the new generation of Learning Power Trackers (LPTs) have to work with this variability, not ignore it.

It might help Jake to fill in his LPT twice, once for his favourite and once for his least favourite subject - or for “school” as against “football training”. Then he can see where his strengths in one domain might be generalised to others.

So far, these low-key questionnaires have mostly been used formatively and informally. Teachers use them to support a sea-change in their own priorities, from focusing on SATs and exams to building their students’

long-term, sustainable learning power.

But it is obviously helpful to be able to use them for comparative purposes, too. How does the average Year 10 student perceive their concentration or their creativity? Do boys differ from girls? And how do these measures correlate with more conventional indicators of school performance?

Working with Patricia Broadfoot and Ruth Deakin-Crick, I have begun to tackle these questions at the University of Bristol. Initial results from what we call the ELLI (Effective Lifelong Learning Inventory) Project, funded by the Lifelong Learning Foundation, are challenging. Learning power seems to correlate only loosely with school achievement: you can do well in exams yet lack imagination; you can possess great tenacity yet fail to achieve. And there are worrying indications that students’ perception of themselves as effective learners declines between key stages 2 and 3.

The growth in the number of teachers who act as “learning coaches” indicates that the idea of tracking the development of learning power has hit a popular vein. It is linked with the notion that young people should be guided into a life of learning. And, as the tail of assessment always wags the dog of learning, the development of LPTs is an urgent task.

Guy Claxton is professor of learning science at the University of Bristol Graduate School of Education. Building Learning Power: helping young people become better learners by Guy Claxton is pound;20 from www.buildinglearningpower.co.ukFor more information about the ELLI Project, contact: Ruth.Deakin-Crick@bristol.ac.uk

Professor Guy Claxton will be a keynote speaker at “Beyond the Exam: innovative approaches to learning and assessment”.

Other speakers include Professor Paul Black of Kings College London and Martin Ripleyof the Qualifications and Curriculum Authority. The conference is organised by Nesta Futurelab at the Watershed Media Centre, Bristol, 19-20 November. Fee: pound;195 plus VAT; day rate pound;130 plus VAT.

Registration: www.nesta futurelab.org