Here, I think out loud about two excellent pieces, one a paper and the other a chapter from a book:
Why Minimal Guidance During Instruction Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem-Based, Experiential, and Inquiry-Based Teaching by Paul A. Kirschner , John Sweller & Richard E. Clark (2006)
Constructivism as educational theory: contingency in learning, and optimally guided instruction by Keith S. Taber (2011) available here.
The current tension between advocates of pupil-initiated inquiry (which researchers like Paul Kirschner describe as ‘discovery learning’) and direct instruction relate to primary science learning. On the one hand, inquiry is at the heart of science learning and without it, it seems hard to see how pupils could understand the nature of science, how theories develop and the role of evidence within this. Yet without disciplinary, ‘canonical’ knowledge and understanding in place, what remains of science? What too of Young’s ‘Powerful knowledge’?
In light of this, how far should teachers use pupil-led inquiry as a means for pupils to ‘discover’ content? (What is the likelihood of this anyway?) How far is inquiry ‘spoiled’ by the teacher ‘telling’ the answers? How far should teachers ‘teach’ the content through direct instruction, and use inquiry, as Popper advocated, to test theory, rather than arrive at it? Would this kind of deductive approach be better suited to the learning of more abstract science content? Would this help to avoid misconceptions that endure long after teaching, and even after extensive science study, as in the Private Universe? Was this teaching disaster the fault of constructivist ‘minimal guidance,’ or just poor pedagogy? I assume Taber might say the latter, and Kirschner et al perhaps both, but it would indeed be hard to say.
Researchers Kirschner et al, group discovery learning, constructivist pedagogy, experiential and problem based learning together and put forward evidence that such an ‘approach’ (if all these can be distilled into one) is less effective for the novice learner. Kirschner et al also describe such approaches as inductivist in that, I’m guessing, observations and findings are considered to lead to generalizations and learning outcomes. Michael Matthews also criticises the constructivist approach to science learning (more on epistemological and ontological grounds) and also describes this approach to science learning as inductivist. (Taber appears to define science inquiry learning as deductive, beginning with the learner’s ‘theory’ as a starting point. Driver and Harlen agree with this, although many of the inquiries Harlen describes in her teaching manuals are based on what appear to be inductive inquiry, as pupils gather data in order to find out.)
Kirschner and friends’ critique of discovery learning relies heavily upon cognitive load theory, which characterises problem based learning (specifically for new learning) as placing unnecessary burdens on working memory, which is limited when processing new information. John Sweller describes problem solving as ‘means- end’ cognitive work that results in working memory capacity being taken up with the ‘process’ rather than building memorable learning. Kirshner and Sweller assert that ‘constructivist’ minimal guidance for new learning is less effective (I agree about the minimal guidance, I’m still not sure that’s necessarily a constructivist approach).
This leads me to this question: is child-led inquiry in primary science an ineffective way to learn new concepts? Is ‘inquiry to find out’ (induction) an approach that uses up cognitive resources so that children are not able to build the necessary ‘canonical’ knowledge and understanding? Would it not be more in line with recent evidence in cog science then, to use direct instruction first, then use inquiry to test new knowledge and understanding, including the role of evidence, when content understanding is intact?
Taber suggests that Kirschner et al have got constructivism wrong (at least Taber’s version of constructivism.) Taber asserts that constructivism does not expound the idea of ‘minimal guidance’, but rather ‘optimal guidance’ based on teachers scaffolding understanding wherever necessary, working with children’s conceptions and facilitating knowledge and understanding in a variety of ways, that might include direct instruction at some point, or exploration at others (which makes sense to me. I also think this doesn’t contradict Kirschner et al). Taber suggests Kirschner et al create an unnecessary ‘either, or’ pedagogical dilemma, asking teachers to decide between direct instruction or discovery learning ( I guess it could be seen this way). Kirschner et al would say (I think) that they are referring to the benefit of direct instruction for the novice learner learning new (‘canonical’) content at the initial stage only, and not as a pedagogical approach to teaching? Taber would agree with the benefits of direct instruction for the novice too it seems, but would add that the teacher might “shift between periods of teacher presentation and exposition, and periods when students engage with a range of individual, and particularly group work, some of which might seem quite open ended. However, even during these periods, the teacher’s role in monitoring and supporting is fundamental,” (Taber 2011:57). For Taber, this is optimally guided instruction. Yet, I’m not sure Taber or Kirscher at al are even disagreeing with each other, as knowing when to drop the direct instruction is as important as knowing when to apply it, surely?
Where does this leave pupil-led science inquiry then? Taber defines extended inquiry as ‘not the most effective way of teaching the focal concepts, but this is still preferred because the primary rationale is to teach students the skills and processes of inquiry” (2011:56)
Are we then to assume that developing science skills is all that really matters? What does this look like without knowledge and understanding in place? Is teaching young children science skills and the process of inquiry the ‘primary rationale’ in science? I’d like to think that it might be sometimes, but not all the time, and certainly not at the cost of Taber’s ‘canonical’ knowledge? Again, perhaps we are (or I am) creating another unnecessary ‘either, or’ pedagogical dilemma. But, if we consider evidence in the field of cognitive science, this cannot be the ‘primary rationale’ for building knowledge and understanding of the ‘canonical’ versions of knowledge and understanding of which Table refers, and which teachers need to scaffold pupils towards. Notably, under my reading, Taber does not make reference to the problem of cognitive load associated with inquiry learning itself, only to the limitations of working memory per se, but perhaps he doesn’t need to as his exposition of ‘optimum levels of direct instruction,’ makes up for this (2011:57).
If, ‘inquiring to find out’ (an inductive process) is a pedagogical device to build skills and develop understanding of the nature of science and the role of evidence, then is there not a place for ‘inquiry to test theory’ (a deductive process) whereby canonical content is taught through exposition (Kirschner at al’s direct instruction) and then tested through inquiry? Would this also not be more suited to abstract, unobservable phenomena that often lack empirical evidence for the learner?
I will pause here for further thought and more reading, suffice to say, there is plenty ahead.
Thank you for reading my on-going and often half-formed thoughts here. I cannot say here that I speak for any of those esteemed researchers I have referred to, but in writing this I have gone some way towards understanding the issues involved for myself. I would urge those interested to read the papers/ book chapter in question as they are both excellent. All comments are welcome, and any mistakes in my references to the research here are entirely mine, and I would welcome corrections.