…from a purely theoretical standpoint alone, it hardly seems plausible that a strategy of inquiry that must necessarily be broad enough to be applicable to a wide range of disciplines and problems can ever have, at the same time, sufficient particular relevance to be helpful in the solution of the specific problem at hand.
David Ausubel

It’s tempting to believe that if we teach children how to think, then they’ll think better. After all, when we teach children to read, then they read better and when we teach them to juggle then they get better at juggling. Why should thinking be any different?

Well, first we have to identify what we mean by thinking. In What Every Teacher Needs To Know About Psychology we say this:

There are two common usages of the term ‘thinking’. One holds that thinking is everything that the conscious mind does. That would include perception, mental arithmetic, remembering a phone number, or conjuring up an image of an elephant headed zebra. We might also include the many varieties of unconscious thought but whilst unconscious cognitive processes may well be tremendously important in shaping the way we make sense of the world, we are using ‘thinking’ in its conscious sense.

Simply equating thinking with any and all conscious cognitive processes is too broad to be useful. Thinking is an essentially active process and therefore distinct from the more passive ‘thought’. Thought is the result of thinking and thinking is the struggle to get from A to B. So, thinking is conscious and it is active. It is the kind of deliberative cognitive process that can make new connections and create meaning. It is dialogic: it has the quality of an internal conversation between different perspectives, although the this is not always immediately obvious. And it is linguistic: verbal for those of us who use spoken language, visual for those of us who use sign language to communicate with others and with ourselves… Thought then, for our purposes, is the inner dialogue we have with ourselves. (p.15-16)

It’s fine if you want to quibble with this definition, but this is what I mean when I think about thinking. So what might a thinking skill be? Depending on who you ask, you get stuff like this:

  • Organising gathered information
  • Forming concepts
  • Linking ideas together
  • Creating, deciding, analysing, evaluating
  • Planning, monitoring and evaluating

Have a look at this website if you feel the need for any more. The thing is, these skills are worthless unless tied to a body of knowledge. In order to organise information you must have some information to organise, but organising information is something we do automatically. Likewise, forming a concept. A concept is formed out of what we know, and again our minds appear to be wired in order to make forming concepts easy. As an intellectual exercise – practising thinking skills if you will – why not work your way through the rest of the list and suggest how any of these items could possibly exist in the absence of propositional knowledge.

The point is we all have an innate capacity to do many of these things already. We’re born with the ability to organise environmental stimuli into schemas which then form concepts etc. We do this unconsciously without the need for thought. Other things like planning and evaluating also happen unconsciously, but we can also decide to pay additional attention when our experience is such that we’re not sure as to outcomes. It might be useful to prompt children to do these things and briefly demonstrate how to do them, but I’d suggest investing much more time than that will run into a considerable opportunity cost.

If you want to, you can take an A Level course in Thinking Skills with Cambridge Assessment. The skills assessed in this course are our old friends problem solving and critical thinking as well as ‘problem analysis and solution’ and  ‘applied reasoning’. Now, of course, you can learn a body of knowledge which includes recognising and identifying biases, questioning assumption and identifying logical fallacies. These are things everyone would probably benefit from learning about and they will, no doubt, increase your crystallised intelligence. The general principle here is, the more you know the better you are able to think.

Now, of course, we can measure raw reasoning power – or fluid intelligence as it’s usually referred to in the scientific literature – but the point that I’ve made over the past few blog posts is that currently, to the best of our knowledge, there’s nothing you can do to increase someone’s fluid intelligence. Here’s the sort of question used to measure fluid intelligence: “If P is true, then Q is true. Q is not true. What, if anything follows.” Some of the questions in the Thinking Skills paper are more insensitive to instruction than others, but that just means, there’s little gain in teaching the thinking skills beyond a certain point. Exposure to, and practise at, these kinds of questions improves your ability at answering them, but ultimately, some people are just better at reasoning than others. 

For instance, how would you teach the thinking skill required to answer this question?


But what about Cognitive Acceleration? Specifically Cognitive Acceleration through Science Education. The claim made by proponents of CASE is that by teaching their science course to 12 years olds, English Language GCSE results were improved at age 16. Too good to true? Well not according to Adey and Shayer, who, in their 1994 book Really Raising Standards documented the miraculous effects of their research. The basis of the CASE intervention requires a mediator to ask questions that allow “guided self-discovery” with students working together in groups to solve a problem.

Clearly, this flies in the face of decades of research in cognitive science and is the ultimate vindication for proponents of discovery learning and group work. Except that in a 2016 RCT funded by the Education Endowment Foundation Adey and Shayer’s findings didn’t seem to replicate. In fact, the conclusions drawn by the research team were that there was “no evidence that Let’s Think Secondary Science improved the science attainment of students by the end of Year 8” and further, “Students who received LTSS [Let’s Think in Secondary Science] did worse than the control group on the English and maths assessments” although they do allow that “this result could have occurred by chance and we are not able to conclude that it was caused by the programme.” Not only was there no effect for Cognitive Acceleration programmes on science, there was certainly no evidence of far transfer.

Despite a significant investment in tailored CPD, “Many schools did not implement the programme as intended by the developer.” There is always a problem with educational interventions that teachers find difficult to deliver. It may be that CASE – or Let’s Think in Secondary Science as it’s now branded – is actually wonderful and the poor results are just the fault of incompetent teachers messing up researchers’ hard work, but equally, it may be that if interventions are so hard to get right then there’s probably a serious problem with their design.

As always, we should return to the words of Carl Sagan: “Extraordinary claims require extraordinary evidence.” If we’re to accept that something as unlikely as minimally guided groupwork in science leads to far transfer years later in unrelated domains then we really need some spectacularly robust evidence.

To conclude: if we were to define ‘thinking skills’ as a body of knowledge that adds to crystallised intelligence and thus makes us better thinkers, then yes, of course we can teach them. But, if we want to suggest that thinking skills are somehow different from our naturally occurring innate capacities for fluid intelligence and also different to knowledge, whether it be of the procedural or propositional variety, then we might be susceptible to the next magic beans salesperson who happens along.

Further reading

  • Greg Ashman has blogged about CASE here and here.
  • It’s also worth reading this paper from André Tricot and John Sweller on why teaching generic skills doesn’t work