Last changed 7 June 1999 ............... Length about 5,000 words (35,000 bytes).
This is a WWW document maintained by Steve Draper, installed at You may copy it. How to refer to it.

Web site logical path: [] [~steve] [this page]

Notes on reflection

Contents (click to jump to a section)

Notes by
Stephen W. Draper
Department of Psychology
University of Glasgow
Glasgow G12 8QQ U.K.

Still to do

Kurt Lewin: do it and learn from (thinking about) the experience.


These began as personal notes on the concept of "reflection" in learning and education; taken from my notes on the Laurillard model of the learning and teaching process. Basically, I have found many different senses of the concept, and these notes try to enumerate those senses. Triggered by the seminar conducted by John Cowan at TLS at Glasgow University on 10 Nov 1998 I revised and expanded the notes. There then followed an email discussion, and that has led to me revising these notes further.

The main messages are:

There are, however, really three issues involved here:

Senses of reflection

[Triggered by the Ray McAleese talk at SRHE'95]
  1. Nearest sense in a dictionary "Conscious thought; meditation". Oxford English Dictionary: "Reflect: go back in thought, meditate or consult with oneself."

  2. An argument with oneself = internalised Laurillard dialogue.
    An interesting question here is whether there is an important contrast between argument, dialogue, and cross-examination. Two themes here being detecting and addressing inconsistencies, and trial and error — that is, producing an action or opinion and then criticising it.
    Re-expression, but do as well as by oneself, so no teacher needed.
    E.g. draw concept maps
    E.g. draw argument structures. Cf. High Perry score, considering the relationships betweeen evidence and several theories.
    => sorting out relationships (among concepts?) by externalisation
    But this is purely at the conceptual level: about relationships between concepts, not experiences.

  3. Laurillard uses it for the steps (by either teacher or learner) of going from experience to concepts to modify the latter. (She uses "adaptation" for modifying action by considering concepts.)
    So concept maps i.e. processing the relationships between concepts would not count.
    * Kolb uses it essentially in this way.

  4. Reflection = self-awareness. I.e. reflexive thinking. True of experiential learning if by self we mean what happened to the self and what it did. Cf. also Suchman, and using plans to describe post hoc what/why you did.

  5. The maintainence of a mental image in consciousness to the exclusion of competing ones = success of one pattern of activation over all others.
    I.e. reflecting one view of the world in the mind. This is a use of "reflection" consistent with the (false) theory that thinking is representation, and representation is mirroring the world in the mind (cf. Rorty). I.e. "reflection" could refer to that subset of thought that is correctly described by those theories of mind and thought.

  6. Reflect as in a mirror; redescribe; use a new metaphor, a new concept to concept connections. E.g. "Reflecting on the idea I heard today". This sounds like (4), but here the mirror is not the mind, but a metaphor got from someone else, or from brainstorming. So it is another technique belonging under (1).

  7. Any kind of processing after the fact; slightly delayed processing of recent events; cf. my theory of stories. If any re-processing by the learner is reflection then it refers to most of the 6 learner activities (4 of them).

  8. Schön reflection: thought about action (results), which may change either action or knowledge. (Reflection in action, reflection on action; knowledge in action, knowledge of action.) I.e. he only applies it to the effects of action. Note too that his main stance is against "technical rationality", but that the Laurillard model tacitly embodies it by having the two level split of concepts and personal action. Schön's point is that you can reflect on and modify action without consulting theory. So Schön's reflection is actually in the iteration within the level of personal experience, and NOT between the levels.
    8.2 Note that for Schön, reflection is about modifying action; and doing so with thinking but without the level of concepts. This is exactly the opposite of Laurillard, who uses it only for modifying concepts.

  9. In education in general, a big concern is to get learners to relate action and concepts. Probably the biggest part of that is from concepts to action i.e. to change the way learners think and act in the world to be consistent with the concepts they have learned.
    9.2 For Schön, reflection uses observations (of effects of action?) to drive reflection. But in my view in education the main effect you want is to go from a new concept to its implications. A big such concern is to connect with your own action; but perhaps even more, is to think through its relationships with other ideas you have (consistency).
    9.3 I would call this "reflection": it is certainly thinking, and not acting, not just going ahead with routine exams and exercises. And it starts from (new) concepts. This is the opposite of both Schön and Laurillard. It is a kind of iteration within the conceptual level; but it is not the Laurillard iteration of just re-expressing it. It is about working out connections and relationships, and may also be relationships to action and personal experience.

  10. Dewey is said to have synonimised: reflective thinking, reflective judgment, problem solving, critical thinking. Dewey,J.(1933) How we think (New York: Heath). And a key attribute of his view of this thinking is that it is associated with problems with real uncertainty.

So ....

So perhaps the common theme is reflect = do more thinking about something (7), making more connections in the mind (or more original inferences); and perhaps actually inferring some new and useful aspect. Output could be two kinds of thing: concepts and plans of action. Two major types of subject for reflection: reflecting about sensorimotor experience, and reflecting about verbal descriptions and concepts.

Hence note that (3) and (4) are mutually consistent, but inconsistent with (2). There are in fact two rather different kinds of connection that we want to prompt learners to make: a) with other concepts and facts they know (2); b) with their own concrete experiences (3,4). These are different things, probably requiring different techniques. Using concept maps is not promoting reflection in the sense of making connections to personal experience. More concretely, they are probably good revision tools, but not the thing for getting students to connect the lab work with the lectures.

In the end, perhaps there are 3 levels of meaning for "reflect" in this educational context:

Synthesis of the kinds of (meaning of) reflection

The only common idea is more processing about something (6,7). It is about thinking, not actual action; and strongly implies no time pressure, but seeking understanding. It also implies no method, but a form of thinking to see what emerges i.e. not strongly goal directed to arrive at a particular end, but rather bottom up driven as a response to an internal estimate of unfinished thinking business.

Currently, I think the varieties of reflection may be spanned using four independent dimensions, some with subdimensions; and almost every combination of alternatives makes sense.

Dimension I. About subject domain or about self

It seems there are two distinct root meanings, each important to some writers on the topic: Thinking about one's own actions bridges both aspects, because one's actions are one important part of one's mental content, and because (better) control of oneself means (better) control of one's actions.

The first basic meaning (A) of "reflection" is to think about something, but to think internally: to brood on stuff already in your mind, not on new incoming material (not new concepts, but revising what you already have some part of; not new data, but new processing of already existing memories or plans for action). One might take it to be an argument with oneself, an internal dialogue as opposed to an interaction with a teacher, a peer, or with part of the external world. As a cognitive function, it might be taken to be about doing maintainence work on internal consistency: checking material against all other material in the mind to detect and attempt to resolve any contradictions.

Dimension II. What is reflected on, w.r.t. what else, when, and for what

Given that, then the key questions are (1) on what material is the thinking done, and (2) when is it done. The first question can be refined: the essence of (type A) reflection is to review or create new connections, so the "on what" question becomes: making connections from what to what? So 3 questions:
  1. What is reflected upon with a view to changing it (e.g. actions, concepts, ...). Kolb and Laurillard reserve the word "reflection" for thinking about and modifying concepts (in the light of experience of action). Laurillard's model however is deeply symmetric, and she uses the term "adaptation" for thinking about (modifying) action in the light of conceptual knowledge. Schön uses "reflection" for both. Some of the main sub-dimensions of this "what" dimension are:

  2. What is it related to or checked against? This question is particularly neglected in many treatments. This in turn may be because the Kolb diagram of reflection depicts only a cycle, and omits the other input(s) that must be there for any act of (type A) reflection. If we took the diagram literally, reflection would mean only the sterile internal rehearsal of the original material, rather than new processing by relating it for the first time to other internal material. That neglect probably is the cause of Colin Holroyd's unease with a tendency to deny the role of fixed, pre-existing, shared knowledge as still important. It also quickly leads you to recognising that there are a large (and open-ended) number of different kinds of reflection, depending on which connections you examine (e.g. check action against theory, against what other theories you know, against what others do i.e. standard practice, ....). Again, the same subdimensions apply, and all combinations can occur e.g. checking a new concept against old private experience, using a new experience to revisit an old (understanding of) a public theory.

  3. When is the reflection done?

    The first form of this question is posed in relation to action events. This is the distinction between reflection for / in / on (depending on whether the thinking is before / during or / after the action being thought about). This distinction expands Schön's in/on distinction, and was made by John Cowan, and is also made on Ray McAleese's home page which has pointers to further material.

    Note however that this does not really cover the important issue of new/old material. For much reflection, one can ask what is the new material that is prompting reflection, as opposed to older material against which it is being processed. Thus reflection-on is largely about taking new observations of actual effects of action and comparing them to the intended effect, while reflection-for is largely about taking a new plan just hypothesised in the mind, and checking it against relevant knowledge (perhaps to catch what Norman calls "mistakes" as opposed to "slips" — plans inconsistent with the actor's knowledge because there was an error in their generation as opposed to execution). But reflection, especially of type A, may be using much older material, sometimes comparing old with very old, as well as comparing a new concept against older ones. Both Galileo and Einstein made important advances in physics by thought experiments: which are reflective exercises that point out to other scientists that there is a conclusion to be drawn simply from reflection upon known material without new data at all.

    Thus this "when" dimension could be recast away from the special question of when in relation to action (which in any case does not apply to all reflection) and towards which, if either, of the two items being related is the new one: which gives three cases to cross with the other dimensions.

  4. What for. In addition is John Cowan's distinction between analytic vs. evaluative reflection. This concerns the regulation of action (without explicit concern with concepts or theories), and so is more strongly related to the type B conception of reflection as to do with self-regulation and understanding through thought. Analytic reflection refers to analysing an action in its own terms, for instance noticing what you did and why, and against all your relevant predictive knowledge (important because circumstances modify most plans, even if you had one, because people sometimes make mistakes in generating even quite ordinary plans, and because many plans are hypotheses to be checked against possible constraints and conflicts). Evaluative reflection refers to comparing what you did and its effect with some standard of what you would wish to have achieved. To do this properly (he suggests), you must describe both the aspiration (standard) and the actual performance separately, and then describe the difference.

    Analytic reflection is formative, to develop abilities; while evaluative is summative, "for deeper and more purposive learning" i.e. to develop the goals and standards you set yourself (not just your skill at achieving whatever you attempt), and again John's claim is that this has a positive effect on learning. For analytic reflection, the new item is the plan of action or possibly observation of behaviour (but not effects), and the old material will be predictive knowledge and any consideration of constraints and aims that an action should conform to. For evaluative reflection, the new item is the effects of action, and the old material likely to be the desired effects. A comparable distinction occurs in Norman's (1986) "theory of action" and its two sides, with action overall being conceived of as comprising first the generation of behaviour, and then perception (as an instrinsic part of action) that is used to observe and evaluate the effects against the desired goal(s).

Note that the first two questions or dimensions are associated with type A reflection, the third may be conceived generally or only in relation to action, and the last only applies to actions and is associated with type B reflection.

Dimension III. Learners or teachers

All of this applies to learners learning about some domain or topic; but it also applies to teachers and all professionals learning while doing their job. So in a teaching and learning situation, there is reflection for and by learners, but also the reflection teachers may do about their own performance (rather than about the subject matter itself).

The relationship of reflection to other ideas about the learning and teaching process

Particularly important classes of object for II.1 and II.2
There are some kinds of connection or relationship that are thought (by various theorists) to be particularly important for learners to make, and so for teachers to support thinking (reflection) about.

Reflection must be a near relative both of critical thinking, and of deep (in contrast to shallow) learning. In fact, we can probably equate reflection with critical thinking turned on one's own ideas and actions as opposed to those presented by others. Reflection is more often used to refer to actions, and critical thinking more often to alternative theories of the same phenomena, but the framework above applies equally to both.

Similarly, there is a very close connection with understanding and deep learning. First of all, recall that deep learners are characterised by aiming not at learning but at understanding (Marton et al.; 1984). Then we must recognise that, in spite of everyday language usage, there is no identifiable completion or end of understanding or deep learning. In everyday chat we may say "yes, I understand", but most serious academics know they continue to pick up new, interesting aspects or consequences of a concept indefinitely (often from their students): aspects they hadn't apprehended before. This is easy to understand in the framework above: there is no definite end to the number of things you might relate a given concept to: complete understanding would mean you had explicitly related it to every other datum and concept in the universe.

Another way of listing types of depth and understanding is given here.

Reflection and the Laurillard model

The Laurillard model to a great extent embodies or includes the Kolb / Schön model of reflection as a part. One of its three underlying principles and symmetries is the distinction between the level of public conceptions and that of personal experience; reflection is seen as the traffic between the two (in fully developed learning processes), where the learner compares, relates the two levels and makes modifications to improve each in the light of the other. Laurillard has 4 of her 12 activities concerned with this, one of which is actually named "reflection". A criticism of her model in the light of the fuller range of "reflection" suggested above would be that her model has no explicit place for reflection (connections) of other kinds, for instance considering the relationship of a new concept with old ones from other topics.

Links between the 2 Laurillard levels

Note the difference between connecting (relating) an item to a) other conceptual items e.g. what was learned last term; b) personal experience, perceptions, etc. And the difference between linking a new concept back to existing knowledge; vs. linking forward by planning actions. I.e. there is the problem that in general it may not just be the new concept and old personal experience that needs to be linked.

Are reflection and adaptation different?

The real issue in reflection in general is thinking over something, checking for new associations and inferences, without having any particular goal in mind: i.e. a goal of understanding, not doing or learning. But the issue here is connecting the two levels of public conceptions vs. personal experience.

Many (most?) people I discuss the Laurillard model with find the adaptation/reflection issues the hardest to see how to address (i.e. to support by explicit teaching actions). E.g. how could you possibly support them by computer? Also, I'm told that young children can't do it at all: this is something only adults can do; and the ability to reflect is a truly developmental one that emerges. Hence Tolmie (personal communication) argued that these 4 activities are qualitatively different from the others. The others are externally visible activities e.g. with material products. Adaptation is too: using concepts to change your procedure, which is then observable. Basically both are observable with one extra step i.e. re-doing one of the first 8 activities.

Examples of support for these 4 activities

A first pass is, that prediction tasks force adaptation, explanation tasks force reflection. Pre-lab activities (such as those developed by Alex Johnstone and others) are directly adaptation tasks (theory to practice). Diagnoses of problems in practicals are directly reflection tasks.

What is important here?

I shall argue that reflection is not worth special attention or emphasis, not because it is unimportant but because, like breathing, although essential for learning it goes without saying. The limiting factors are elsewhere. Because of the vagueness of the term "reflection" (i.e. the inconsistency of its usage) the point needs to be made for each major meaning. I shall therefore discuss separately reflection by teachers and reflection by learners; and consider for each three meanings: thinking, iteration, and relating the two levels of public concepts and personal experience.

Reflection for teachers as professionals


The base (dictionary) meaning of reflection is thinking. Yes thinking is important, but doesn't that go without saying?


A deeper idea is that of an iterative cycle, as discussed by Kolb and Schön. The idea of learning and developing by acting, extracting lessons from the experience, and so improving one's capability for that kind of action.

Yes, that is a profoundly important and general concept. It is also old. It is the key idea in cybernetics (and the influence of that on what is now cognitive science many decades ago): the idea of applying the concept of a feedback loop from control engineering to human behaviour. It is the idea in the TOTE cycle (Miller et al. 1960). It is the idea in Don Norman's Theory of Action (Norman, 1986). And it is represented in the Laurillard model by the iterative groups of activities and loopbacks between them e.g. activities 1-4 (instead of just 1 and 2) etc. In other words, we haven't waited for the notion of reflection to hear about the need for such cycles, which are much wider significance. In fact, they seem basic to all action, as well as all learning.

On the other hand, in my experience of teachers, they haven't needed to be told to reflect in order to notice problems with their activities and to consider changes: neither thinking nor expecting a cycle seems a problem. Instead I have seen two different problems or effects.

The first is in reporting back to teachers with an evaluation report on some piece of their teaching. Typically they don't need suggestions or even interpretations, but seize on the data and immediately think of changes they will try out. This suggests that a bottleneck was getting good feedback from students about the activity. Thus techniques like ones John Cowan mentioned are the key (e.g. starting a lesson by asking them to write down what they most want to learn in the next hour; ending by asking them to write down the most important thing they learned, and the most important thing they still haven't learned). Most teachers I know will react to this information without any urging to "reflect!", but they don't normally get such information. Providing it is what makes the difference.

The other observation, this time backed by the literature on the ineffectiveness of standard course feedback questionnaires in stimulating change, is that besides useful data, teachers sometimes need new ideas about what changes they could make: ideas about alternative practices. This too is not supplied by reflection, but by having contact with good ideas about teaching, usually from other people.

Relating public concepts and personal experience

The last observation suggests again, not that teachers need to be urged to do this kind of thinking, but that the bottleneck, at least in HE, is usually having access to relevant abstract conceptions about teaching.

Reflection for learners (to be organised by teachers)


Again, who could dispute that thinking is good for learning? Indeed, although some practices (e.g. lectures) may tend to suppress thinking by learners, many others are, and have always been, specifically designed to provide it: exercises, essays, etc.


Who could need reminding of its importance? Advocating iteration between action and reflecting on its results amounts to advocating giving students feedback on their exercises. Iteration in general is an underlying principle in the Laurillard model represented by the "to and fro" between left and right sides of the diagram, and by having four rather than two activities in both top and bottom halves of it.

Relating public concepts and personal experience

As discussed earlier, this is important and a feature of the Laurillard model. It is also only one special case of all the kinds of thinking and relating that are desirable.

In all cases, the real bottleneck for learners seems to be time for reflective thinking, not being urged to do so. Certainly that has been true for me from the age of 16, when I finished doing a physics course in a slow stream for those supposed not to be interested, but which thus left me time really to think about each topic in depth. After that, rushing greatly reduced how much I could think through and therefore understand what I was taught, until my PhD.

Learners can't think because teachers put them under endless pressure to meet deadlines and push on to the next topic. And the same is seen everywhere. One of the very irritating things about (British) TV is the way interviewers constantly either interrupt or put the next question within a split second of the last reply. They never ever stop to think about the answer, nor allow their audience to do so. This continual demonstration of disinterest in what is said is rude, and prevents understanding, thought, and reflection.

Similarly as a teacher, what I probably need most to be continually reminded to do is to allow silences for thought. Calling this "reflection" is fine, although "thought" is plainer English.

Why reflection isn't worth focussing on

It is far from clear (to me) that reflection is worth much prominence in either theories of the learning and teaching process, or in our teaching practice. I feel it is a secondary not primary issue. This impulse to be dismissive has these (not necessarily mutually consistent) components:

  1. I find there are many senses of "reflection", and most who use it don't seem to have reflected on this. If new practice is worth developing there should be a clear theoretical position behind it; and if there is a clear theoretical position then at the very least clear definitions and discussions of how this is and is not consistent with other usages should be available. This is not the case, as I hope the above makes clear.

  2. This unclarity means among other things that perhaps we all do it anyway, like the man who was astonished to discover that he had been speaking prose all his life; in which case most of us need pay no new attention to it. After all, "reflection" just means "thinking", and I'm sure we all agree that thinking is helpful to learning.

  3. The same issue occurs in other theories (such as the theory of action: Norman, 1986). And there, it is conceived of as an inherent part of action (like breathing), not some new discovery to be promoted by special approaches because otherwise ordinary learners and practitioners will omit it.

  4. It is already an assimilated (unemphasised) component of the more interesting theories in education, such as Laurillard's and the work on deep and shallow learning. Making a fuss about reflection is to focus on one component and ignore the overall picture.

  5. Similarly for practice: reflection is a part of or effect of good practice, and there is no obvious need to focus on it more than, or separately from, others. It will occur without special attention. For instance, if the Laurillard model is applied systematically to the design of teaching provision, reflection will have its place, just as exposition will.

  6. As for promoting reflection for learners, which is reflection in the sense of thinking about oneself, we should remember that shallow learning is characterised by striving to learn: perhaps promoting learning diaries etc. will erode deep learning and promote shallow learning. (Understanding diaries are what that work would predict are required.)

  7. In considering reflection for teaching practitioners, then if your aim, like mine, is to improve the quality of L&T in HE, then in my experience it is not reflection that is usually the bottleneck. Instead, one bottleneck is getting useful feedback on my teaching: so techniques like ones John Cowan mentioned are the key (e.g. starting a lesson by asking them to write down what they most want to learn in the next hour; ending by asking them to write down the most important thing they learned, and the most important thing they still haven't learned). Most teachers I know will react to this information without any urging to "reflect!", but they don't normally get such information. Providing it is what makes the difference. The other important bottleneck is having ideas about what new practices to try, given information on problems in L&T. This too is not supplied by reflection, but by having contact with good ideas about teaching, usually from other people.

So: Reflection is a secondary issue, like vitamin B. We all need vitamin B, but it is only one among a large set of necessary elements. It is not important like the general concept of vitamins. And when you analyse it, you find there are many different vitamins all separately essential: but originally confounded as one "vitamin B".

But, against that:

But, against all that, is the witness of some enthusiasts that say it benefits them personally. And that experiential evidence could be worth more than all this analysis.


Boud,D., Keogh,R. & Walker,D. (1985) (eds.) Reflection: turning experience into learning (Kogan Page)

Boud,D., Cohen,R. & Walker,D. (1993) (eds.) Using experience for learning (Open University Press)

Dewey,J.(1933) How we think (New York: Heath)

Dewey,J.(1938) Experience and education

Kolb, D.A. (1984) Experiential learning Prentice-Hall Inc, Englewood Cliffs, NJ, USA

[not checked] Kolb D A (1976) The Learning Style Inventory : Technical Manual McBer & Co, Boston

Kolb, D.A. & Fry,R. (1975) "Towards an applied theory of experiential learning" ch.3 pp.33-57 in Cooper,C.L. (Ed.) Theories of Group Processes, (Wiley: London)

Laurillard, D. (1993) Rethinking university teaching: A framework for the effective use of educational technology (Routledge: London). See also an online diagram of her model

Marton,F., D.Hounsell & N.Entwistle (1984) (eds.) The experience of learning (Edinburgh: Scottish academic press)

Miller G.A., Galanter E., & Pribram K. (1960) Plans and the structure of behavior (New York: Holt, Rinehart & Winston).

Norman D.A. (1986) "Cognitive Engineering" in D.A.Norman & S.W.Draper (eds.) ch.3 pp.31-61 User Centered System Design (Erlbaum: London).

Perry, W.G. (1968/70) Forms of intellectual and ethical development in the college years (New York: Holt, Rhinehart and Winston)

Rorty, R. (1980) Philosophy and the mirror of nature (Blackwell: Oxford).

Schön, D.A. (1983) The reflective practitioner: How professionals think in action (Temple Smith: London) (Basic books?)

Schön,D.A. (1987) Educating the reflective practitioner

Schön,D.A. (1991) (ed.) The reflective turn: case studies in and on educational practice

Suchman, L.A. (1987) Plans and situated actions: the problem of human-machine communication (Cambridge University Press: Cambridge, UK)

Web site logical path: [] [~steve] [this page]
[Top of this page]