Practical: The Fallacy of Induction

@SotonEdThis is the third and final post in a series about the value of practical work in science. In the first post I have suggested that science trainee teachers (and possibly some qualified teachers too), have a tendency to make assumptions about the value, and the learning, associated with practical work in science. In the second post I illustrated this with an example and briefly tackled two questions I think are important: whether or not children enjoying practical work is sufficient justification, and whether or not just doing practical will make them better at it. I left a third question hanging and ask it again now.

Do children learn important science ideas and/or develop their understanding from seeing the theory ‘in the flesh’? Often trainee teachers think that this is self-evident. I’m not convinced.

Some of the most useful work on children’s ideas, and misconceptions in science was completed by Rosalind Driver and colleagues in the 1990s. I think this is an essential resource for all science teachers because of the evidence that knowledge of children’s misconceptions is an important distinguishing feature between more and less effective teachers. Some may find elements of the suggestions for classroom practice overly constructivist but for me, as well as the identification of a whole range of misconceptions, the other really useful idea I have taken from this work is the ‘fallacy of induction’.
The fallacy of induction is the mistaken belief that children, when presented with relevant evidence, for example from practical work, will tend to work out (induce) the appropriate scientific theory.

The problem is that correct scientific theories are often simple when you know them, but are tremendously hard to generate directly from evidence. After all, it took a lot of very skilled scientific thinkers hundreds of years to do this the first time. What’s worse, children inevitably develop naïve theories as they grow up, so in secondary school they are sometimes not just trying to learn correct scientific thinking but are trying to un-learn naïve thinking that serves them perfectly well outside the classroom.

As teachers, we can of course select learning experiences, including practical work, that provide far more scaffolding and direction than Galileo, Copernicus, Newton, Darwin, Lavoisier, Faraday or Wegener were working with but, however well we do this, I think that induction from practical work, or other activities, is doomed to failure.

For conceptually straightforward science, where there are no misconceptions to overcome, I think that, as a science teacher, we can clearly see how the correct scientific principle follows from the practical observations and it is very easy to assume this will be apparent to the learners too. For the teacher, the scientific principle already exists as a complete and correct schema (like a mind map) in their long-term memory, and they know which features of the practical are relevant, so making this match is relatively easy. For the learner this is not the case. They just don’t have enough of the necessary knowledge chunked in long-term memory to manage the cognitive load – they can’t see the wood for the trees. Like many cognitive load problems, it may be possible to scaffold or adapt the activity sufficiently to allow children to see the wood, but you have to question whether a forest is the right starting place, or whether a nice piece of rough sawn timber from B&Q might be a better option.

Where there are misconceptions, Driver and others have suggested that cognitive conflict, created by exposure to direct evidence that the existing ideas are untenable, will help to resolve the problem. That was certainly my thinking for many years. It seems obvious that, when presented with evidence that is in conflict with their misconceptions, learners will tend to respond by correcting their ideas (their mental representations or schemas). What actually seems to happen a lot of the time is that they ignore, fail to focus on, or distort the evidence, so that their naïve theory survives and may even be reinforced. This explains why so many intelligent people stuck with Aristotle’s ideas about force and motion for a thousand years despite blatant evidence to the contrary.

The ideas of Daniel Kahneman and others help to explain why people have an overwhelming tendency to respond in this way. David Didau in his #WrongBook is also very good on the reasons why our response to contradictory evidence tends to be irrational.

My personal experience is that I have eventually learned the situations where my quick thinking will be wrong and I need to over-write with the correct scientific idea. For something like weight and mass I can pretty much do this automatically but with something more taxing like the tendency to see speed and mistakenly think as if force and acceleration behave in the same way, the best I can do is stop myself and know that I need to think very hard and apply Newton’s Laws with great care.

I don’t think typical practical work ever produces enough clarity in either the results or the conclusions to even begin to address these stubborn misconceptions. I love asking hinge questions, like the Veritasium videos, that throw up misconceptions, but the next step is to tackle the problem head on. I don’t think there are many situations where children can discover scientific principles directly through practical work and I think it even less likely that misconceptions can be effectively challenged and addressed.

So, what role does that leave for practical work in teaching science? I think, if you’ve read this far, you might be thinking there isn’t much practical work in my science teaching, and that perhaps the children taught by my @SotonEd trainee teachers aren’t getting much either, and what little they are getting is restricted to training in purely practical skills – accurate measuring, and manipulation of equipment. Not so! For me, practical work is terrific for the stage beyond basic theoretical knowledge, for three reasons:

Science is stuffed with abstract concepts and there is good evidence that concrete representations help children to understand these abstract concepts. I think sometimes physical models are more useful but practical work can often play this role. For example, you can find a good, clear explanation (with diagrams and perhaps photographs) of chromatography in any textbook but I think the actual physical process of separating out ‘black’ ink colours makes a big difference to children’s grasp of what this really looks like, and the time scale – that painfully slow diffusion – over which it happens.

Secondly, when new knowledge is acquired it will be very fixed to the original context. Deeper understanding comes from making this knowledge more flexible and filtering out the key points from the peripheral detail. Practical work provides an excellent additional level of complexity through which the scientific principle can be seen. Another way to think of this is that children often need to encounter the same idea in several different ways before it sticks; again, a practical can provide this.

Finally, there is something joyful about seeing abstract theory writ large (or often actually quite small) in the fabric of the universe. Science differs from other subjects because it is humankind’s ultimate attempt to describe, and perhaps even understand, the physical world around us. As science teachers, we need to be careful not to think that children see practical work the way we do, but if we ever lose the joy then it’s time to do something else.

Driver R. (1994). The fallacy of induction in science teaching. Chapter 3 in Levinson, R. ed. (1994) Teaching Science. London: Routledge

Nuthall G. (2007) The Hidden Lives of Learners. Wellington: NZCER Press

Pashler H., Bain P.M., Bottge B.A., Graesser A., Koedinger K, McDaniel M and Metcalfe J. (2007) Organizing Instruction and Study to Improve Student Learning: IES Practice Guide. Washington, DC: National Center for Education Research, Institute of Education Sciences, U.S. Department of Education

Sadler P.M. and Sonnert G (2016) Understanding Misconceptions: Teaching and Learning in Middle School Physical Science. American Educator. 2016 (Spring)

Shtulman A. and Valcarcel J. (2012) Scientific knowledge suppresses but does not supplant earlier intuitions. Cognition. 124(2) pp.209-215

Thorn C.J., Bissinger K., Thorn S. and Bogner F.X. (2016) “Trees Live on Soil and Sunshine!”: Coexistence of Scientific and Alternative Conception of Tree Assimilation. PLoS ONE. 11(1)

Willingham D. (2002) Inflexible Knowledge: The First Step to Expertise. American Educator. 2002 (Winter)


Practical: Grinding Frustration

I’ve seen a lot of trainee teachers knacker lessons up with a well run, but ultimately pointless, practical. Whole-class practicals, in particular, are massively time-consuming with many filling an entire lesson. If all that has been learned in that time is “It went blue, sir” then I don’t think that’s good enough.

The problem is usually a confusion over learning objectives. My previous blog set out the way I see learning objectives in relation to practical work but I’ll recount an example. I went to visit a promising trainee teacher a few years ago; the lesson was part of a unit on separation techniques I think; this was certainly part of a sequence on chromatography. When I looked at the lesson plan, saw that it was mainly going to be chromotography of pigments from leaves, and that the learning objective was “To separate leaf pigments by chromatography” I tried to help by asking her what she actually wanted the children to learn. I just couldn’t get past “I want them to separate the yellow pigment from the green chlorophyll”. It’s not fair to pull the rug just before an observation so I let it go and waited to see. Bless those lovely Y8s; they chopped and crushed and ground their little hearts out. They followed the instructions as well as they could, set up their chromatography paper (several submerged the spot in the propanone), and then did a little write up whilst they waited for the chromatograms to be ready. Some got a bit of green and yellow differentiation and the rest didn’t. Whilst they were working I went round and asked a few questions, such as “Can you tell me why you’re doing this?” and “What’s the point of chromatography?” I didn’t even get half answers, just pretty much universal “Don’t know”.

In the feedback session I didn’t get any further, really. The trainee teacher was very disappointed with the lesson. She carefully evaluated the quality of the practical work and made some perceptive comments about maybe splitting the practical into sections and briefing more closely to ensure the leaves were finely chopped, the amount of propanone was reduced, and the papers were set up correctly. But she completely and stubbornly failed to identify the problem, which was that, her ‘learning objective’ wasn’t about learning at all; it was about getting the practical to work. Had the chromatograms come out well, she would have been satisfied with the lesson,  Even when I directly asked the questions “Did the children understand the process?” and “Did the children understand what chromatography was?” and pointed out that they had been unable to tell me anything about these things, she couldn’t really see that this was a much bigger problem than the poor results.

There are plenty more examples where that comes from. Some worked nicely as expected. Some didn’t. All suffered irrevocably from a sense at the planning stage that the practical somehow justified its own existence just by being practical. Often, I find a defensiveness of practical work that I don’t see when pointing out other misaligned learning objectives. That sense that practical is self-justifying can be difficult to change. Why is this difficult?

In the end this boils down to the questions of whether or not (a) children enjoying practical work is sufficient justification, (b) just doing practical will make them better at it, and (c) children will learn important science ideas and/or develop their understanding from seeing the theory ‘in the flesh’. Often I think trainee teachers think, perhaps sub-consciously, that some or all of these are self-evident. I’ll tackle (a) and (b) here and leave (c) to another blog.

For me, enjoyment can’t ever be an end in itself for what happens in science lessons; that just reflects my personal belief in what school is for – no evidence presented. On the other hand, if enjoyment leads to better learning, higher motivation, more time doing science outside lessons, improved post-16 take up, and so on, then the judgement is maybe about balance between enjoyment and learning. I don’t have the expertise to offer a definitive review of the evidence but I’ve certainly been influenced by Abrahams (2009) Does Practical Work Really Motivate? and I’m not convinced practical work is as critical to motivation as is often assumed. The ASPIRES final report makes a brief reference to reduced interest in science after Y9, which might or might not correlate with reduced practical; personally I think it is the  GCSE curriculum content, and looming exams, that is to blame, but can’t offer more than a hunch.

Is it good teaching to explain how to do something tricky and complicated, and then get the children to try lots of examples with very general (that one’s good, that one’s bad) feedback? No, of course not. So why would practical skills be any different? Most of us have had years and years of experience through school, university, and maybe in the classroom, to hone our practical skills. Many of us have probably also taken things to bits and re-built them, developed fine motor, and problem-solving skills, through art and craft and cooking and all sorts. We tend to massively underestimate how difficult it is to extract chlorophyll from leaves, prepare a cheek cell slide, or connect up lamps in parallel. The cognitive load for these things, for children, is very high. In the lesson described above, the instruction sheet and the teacher were both clear about the level of the propanone on the chromatography paper, but at least a third of the class submerged the spot. There was just too much new information for them. These things need breaking down, step by step, with practice or at least immediate feedback at each stage. Without this, children just get used to practicals not working half the time (and working more often for the ‘smart’ kids and more rarely for the others) and accept this is the way of the world. Sometimes there is value in unexpected results, but not if a shrug of the shoulders is the typical response. If we are trying to teach practical skills then we need to plan carefully for those skills, and get precise and accurate work from the children.

Which takes me back to that chromatography lesson. I would have been very happy if the learning objective had been something like “To improve practical skills: use of mortar and pestle to extract plant material; setting up chromatograms; precise working” and then the trainee teacher’s reflection would have been at least a useful starting point. That was an aspect of the intention, but actually, if I’m being generous and assuming the practical wasn’t just picked because it was on the SoW, the stronger intention was something vague about understanding chromatography better by doing a practical example. Failure to separate learning practical skills from developing understanding is a big problem but this idea that doing a practical will improve understanding is, I think, the worst mistake.

Next blog coming up…




Practical: Young people’s view on science education from the Wellcome Trust Science Education Tracker

This morning the TES published a confusing article on key findings from the Wellcome Trust Science Education Tracker. This is a survey of over 4000 young people Y10-Y13 asking about their views on their science education and careers. The TES don’t even seem to have managed a link but the tracker, including a breakdown of the questions and responses is at

Hopefully readers from the science education community might have quickly got past the ‘hands thrown up in horror’ headline and be asking whether the survey tells us anything useful about the quantity or quality of practical work in schools and colleges. Actually there are 144 questions and only 3 are about practical work. There is a mine of useful data here for questions around post-16 STEM participation, science capital, and availability and participation in triple GCSE, which has been a problematical issue, but that’s probably best seen through the lens of the ASPIRES2 work. Hopefully they’ll blog about the survey results at some point.

I’ve only had a quick look but these are my first impressions of the 3 questions (T66-T68) directly asking about practical work.

Firstly, some caution is always required when dealing with self-report measures, and also the way the responses are reported. For example (T66), young people might well have different views on what constitutes “Designing and carrying out an experiment / investigation” and “A practical project lasting more than one lesson” but I can’t see how any of last year’s Y10 or Y11 could not have completed an ISA across multiple lessons. The responses to these two questions were about 75% and 55% respectively with 10% responding “None of these”. What were the 10% doing? Did at least 35% squeeze an ISA into one lesson or do their ISAs in only one of the two KS4 years? How many didn’t think an ISA was an investigation (justifiably?). My take on this is that we need the responses to the same question for current Y10 to see the impact of the new GCSEs, otherwise we are discussing history, but I’m not convinced about the merits of practical projects and multiple-lesson investigations anyway.

Secondly, it’s important to interpret the findings critically. About 1/3 were happy with the amount of practical work and nearly 2/3 would have liked more. As pointed out in @alomshaha’s excellent video, this might be because practical is an easy option, not because it is the best way to improve learning. Even children have a keen awareness of this issue; in the Student Review of the Science Curriculum (Murray & Reiss 2003), about 70% had “Doing an experiment in class” in the top 3 most enjoyable activities (along with watching a video and going on a trip) but only about 40% thought it top 3 for “Most useful and effective activities”.

However, there is one thing in the data we ought to be thinking about. These are the figures for “When doing practical work, how often would you say that you just followed the instructions without understanding the purpose of the work?”


That suggests this statement is true for maybe 1/3 of practicals; this concurs with a lot of practice I see out in schools (from trainee teachers, mostly, but I have a suspicion it’s quite widespread). I think this is a problem.

It’s really good to see a School Science Review article by Millar & Abrahams (2009) here on the AQA website. This is a summary of a significant bit of work they, and some others, did looking at the effectiveness of practical work. Essentially the problem they identify is confusion over learning objectives. Just like all lesson planning, the objectives need to drive the activities, not the other way round. Whole-class practicals form such a big and obvious chunk of a lesson that it’s really easy to start planning from the activity. The trouble is that you then lose sight of the wood for the trees so that a successful practical outcome becomes the real objective – the one you focus on – although the lesson actually has an objective related to knowledge and application of science content. You then emphasise the procedure and just hope the children understand how it relates to the science content. And the children then just follow your instructions (hence the survey response) and, as Millar and Abrahams put it the emphasis becomes “producing the phenomenon”.

Millar and Abrahams go on to suggest there are three broad categories of learning objectives that are served by practical work and, based on a related article, I’ve broken these down further. I find this really helpful in getting a clearer focus on what purpose the practical serves in the lesson and therefore what the best way to approach it is.


If conceptual understanding is what you want, then the children need to spend time thinking about the practical in relation to the relevant content. There are maybe three options here:

  • Whole class practical with lots of time afterwards to do work on how the practical is a demonstration of the content.
  • Whole class practical with very high level of practical competence so children have capacity to think about the content.
  • Demo or video (maybe simulation) so children don’t have to think about manipulating equipment and the teacher can direct their attention with questions and explanations.

The second of these could come from prior learning, but could also be a result of very careful briefing. This is, I think, what @oliviaparisdyer is describing in her blog post about practical work. It is certainly how I remember my excellent O-Grade Chemistry teacher doing it, several decades back into the last century.

If investigative skills are what you want then don’t try to teach conceptual understanding at the same time and remember that as science graduates we tend to massively underestimate the complexity of designing and conducting a full investigation. That’s why the ISAs were such an unpleasant exercise in trying to temporarily get children to remember enough to hit whatever ridiculous coursework target grade they had. I’ve had the good fortune to work with A-Level students on some terrific independent projects (for A-Level Physics and EPQ) but even post-16 they are barely ready for high-quality work. In my experience, either very high-levels of scaffolding, or acceptance of interesting but very rickety work, are needed for 11-16 classes, though that may not be true for all teachers.

Finally, if practical skills are what you want, then again you need to focus on them. Something like reaction of copper(II)oxide with sulfuric acid and then filtering and evaporating to get copper(II)sulfate involves a stack of excellent practical skills to do well. This would be a great practical for improving these skills; I think it’s a massive waste of time for learning the chemistry of metal oxide + acid reactions. By all means combine the two, so do the practical in that unit, and start or finish with the chemistry, but don’t expect the children to learn anything about the chemistry content whilst trying not to scald or gas themselves or – more hopefully – produce nice blue crystals.

This blog is already a bit long; next post I’ll try to use an example to explore these ideas about confused objectives a bit further, and then I’ll try and write another post on why children don’t automatically develop understanding from seeing a scientific principle ‘in the flesh’ and about Driver’s excellent Fallacy of Induction.