Last changed 9 Dec 1997 ............... Length about 5,000 words (27,000 bytes).
This is a WWW document by Steve Draper, installed at You may copy it. How to refer to it.

Learner confidence and motivation

This document is an essay representing my current ideas on this topic, synthesised mainly from discussions with others. The biggest contribution comes from a discussion on itforum.

I made my own edited version of the original discussion. This version has been re-ordered a bit and given a contents list. It was originally created to offer an immediate record, but now the regular itforum web version (which also has an index) is probably more historically accurate.

THIS document is being reconstructed; it is not coherent; it may not be finished till January.

Junk Contents

Preface /Introduction

This essay is about issues of learners' confidence, which turned out to be closely connected to the topic of motivation.

Some introductory commonsense.
Must have motivation: he learning depends upon conscious and actions and effort.

Signs of a problem: the phenomenology

This section aims to do two things together: to say what the problem is, what the warning sings or puzzles were that lead, or at any rate led me, to look for understanding; and to describe (briefly) what each problem feels like from inside. Knowing what something feels like is part of understanding it.

my students rationality paradox baird bev

the nutter

It has been argued that giving learners confidence was all that matters.

Student feedback

In a focus group with students asking them about the feedback they were getting on a course (in tutorials on their programming exercises) they said that what mattered to them was getting some confidence from the tutor. This makes a nonsense of my concepts about levels of feedback, and indeed about all the theories of teaching and learning I usually attend to.

The question for me came from a context of students learning how to program computers on a conversion MSc course, where a) there is a big range of prior ability, so I see many students who don't need any confidence boosting, but also many that do. b) programming gives the intrinsic feedback of whether the program runs which seems to me (from of course a perspective of confidence) to be enough.

Your answers give me a somewhat better grasp on that difference. For me part of the puzzle is or was that by "motivation" I usually mean a decision on how to spend one's time e.g. if coursework is for credit, then effort, in not then none. And these students do not lack motivation in the sense of desire for success and willingness to make an effort. So it isn't obvious to me that confidence is at all a related issue. But some of the reminders you sent it in fact are making a connection across that gap from deciding I could do it but don't think it's worth my time, to deciding that it's not worth my time because I couldn't do it successfully.


I have failed to learn, therefore I feel I don't want to learn.


I will avoid trying to learn, because if I then fail it will be evidence that I am incompetent at learning i.e. failure to try is less painful than failure to learn after trying.


Some theories of motivation define it in terms of behaviour. This leads to some consequences that sound silly e.g. you are tempted to steal a million pounds you find lying in a bag, but instead hand it into the police. This would be described as you not having enough motivation to be rich, as would failing to claim a million that is rightfully your because you can't be bothered to fill in a claim form. Or in a learning context, such theories would equate someone who failed an exam because they lay around watching TV instead of studying, with someone who was prevented from studying by having to care for a sick relative.


On the other hand, many people respond to having a bad time learning something (e.g. mathematics) by ceasing to wish to learn it. This is a real psychological phenomena that happens frequently (but not always), yet is essentially irrational: confounding how desirable something is with how hard it is to get.



Initial solutions

Must have motivation: HE effort. Both value and feasible cost: multiply / AND. Affective LObjs.

A synthesis

Types of 'confidence'

The meanings of "confidence" I identify are:

Three more types of confidence

[taken from a message by Brian Whitworth]

The review I have done suggests there are three types of confidence:

  1. Individual confidence based on intellectual analysis - "My thinking convinces me this is true" not as common as one might think, since how many of us are truly independent thinkers?
  2. Confidence which is conveyed from another person whom we trust and believe in - such as a teacher. "I believe this because the teacher said so, and I trust them". This requires a person-to-person interaction.
  3. Confidence arising from holding the same position as a group we identify with. "Everyone believes this, and so do I". When my subjects found that they were in agreement with the group, confidence went up, even though the computer-mediated interaction was anonymous, so personal trust could not be a cause.

Types of 'motivation'


There are basically two classes of theory of "motivation". There are theories that define motivation in terms of behaviour or need alone (e.g. Maslow): these theories only talk about the net result of mental calculations. The other class (cognitive or process theories e.g. Vroom) talks separately about the strengths of desire or reason for and against acting.


A rationalistic theory would see an agent, such as a person, calculating both costs and benefits and probability of success. In HE (higher education), all three seem to be important variables in practice: learning takes effort, a considerable part of that effort does not depend directly on the teaching, but on time and effort put in out of class by the learner. But there is no certainty of success: in most courses, some fail thus wasting all their effort and other costs. [not just prob. benefits are certain, success is not.]

Motivation and learning:
Steve's synthesis


The exchange in itforum above was notionally on the topic of learner confidence. One of the major links pointed out was the relationship of learner confidence to "motivation". Here is a provisional personal synthesis on the issue of motivation and learning, based partly on some earlier notes of mine, but also on a side discussion I had at this time with Bev Taylor. Bev is just completing a PhD on learner motivation and an attempt to construct an intervention that modifies it. On the measurement side, she used a questionnaire instrument to measure levels and types of learner motivation, the AMS devised by Vallerand (see references).


There are several distinct issues here:
  1. General classification of motivation types
  2. Need to look at costs as well as benefits. I.e. most motivation is about beneficial outcomes, but fails to treat costs separately. Yet a lot of the issue is the amount of effort and whether this is sufficient.
  3. Time scales. Things that persist, things that can be instantly changed.
  4. Experimental demand characteristics that may change the motivation e.g. Hawthorne effect.

General classification of "motivation" types

Types of motivation

The most important distinction to make in types of motivation is between extrinsic motivation and intrinsic motivation. E.g. wanting to pass an exam or to please one's family are extrinisic motivations to learn, and wanting to know why the sky is blue for its own sake in an intrinsic one.


I would subdivide extrinsic motivation into desiring the goal of having learned from willingness to make an unpleasant effort. Vallerand and others subdivide extrinsic motivation (differently) into external, introjected, identified and integrated regulation. These are increasing degrees of self-determination and of integrating the motivation into one's own other goals and attitudes.


I would subdivide intrinsic motivation into prior interests e.g. gold attracts most people in a museum exhibition, and links to personal experience. Whatever the topic, linking it to the learner's personal experience will almost always increase the interest (i.e. the intrinsic motivation) on the spot.

Museums and magazines show how important intrinsic interest is. In a different subdivision, interest is a function of connections to the already known and surprise. So the greater the number of connections of the new item to existing knowledge, the greater the interest; but surprise can only occur when you already have a predictive theory of that item/event. Enthusiasts will examine lots of data, probably looking for and building patterns. So the final theory is, it is the relation of the new item to rules. The more rules you have, the better the chance new items have of testing them; and perhaps scoring highly as interesting by disconfirming them; or equally, allowing a new pattern to be noticed. So perhaps the pleasure of interest depends on new rule formation.


In my view the self-efficacy and learned helplessness literature show a third hybrid type between intrinsic and extrinsic motivation. Basically they are to do with how experience and the attitudes derived from personal experience affect the desire to act: if you don't think making an effort leads to success at this goal, then desire for it and willingness to make an effort both go down. While the latter is rational, the former is not: but it seems our minds do both together.

A different hybrid is used by Vallerand and others, called "amotivation". This apparently refers to some impulse to do an activity (e.g. attend a class) without any belief that this will lead to a desired outcome for the learner of any kind. This is important as discussed below: basically because a necessary (and so important) aspect of learning is a degree of trying and learning things before you know why they are worth knowing: trusting teachers or other people, attending to what you see others attending to, before you know why they are attending to them. This is learning by imitation, but at the level of attention or motivation (perhaps the only or main level at which it can work). Put another way, the motivation is not to learn but to do the surface activities that in others may be done as instrumental in learning (e.g. attending a class), although in other cases (such as chatting to people at a party), it may not be normally conceived of as related to learning, even though learning often occurs as a side effect. On the one hand we must expect reduced learning because the motivation does not extend to the mental processing required for learning, but on the other hand we should recognise that learning can and even must sometimes occur without such intentions but simply as a consequence of surface actions undertaken for other motives.

Deep and shallow learning

Some may think there is a close association between deep and shallow learning on the one hand, and intrinsic and extrinsic motivation on the other. In my view there is no systematic or necessary association. Shallow learning is characterised by focussing on what the test task is expected to be, and learning only what is needed for that. Because the test is external, you might think this has to be extrinsic motivation. However someone might want to learn first aid for its own sake, not because this brings any external reward, but because their own values tell them this is desirable. Nevertheless their learning will be focussed on a specific test task, not on understanding medicine. Deep learning, in my interpretation of the literature, is focussed not on learning at all but on a desire to understand. Although that desire may usually be due to intrinsic motivations, it is quite possible to imagine extrinsic ones. For instance, one day your child comes home looking terrified and refusing to talk about it. You want to understand, not to pass some specific test such as getting your child to talk. Not understanding would be dangerous, you think, but you do not know what it is you have to learn, nor do you have any intrinsic interest about child psychology for its own sake. This would be extrinsically motivated deep learning. (I think this is quite general: understanding may be driven by an extrinsic fear of something going on that should be investigated, without any prior knowledge of what it is that is to be uncovered.)

Basically, motivation is about the will and choosing to act. Learning is partly about involuntary effects (some learning is involuntary e.g. do you remember talking to your friend today? did you have to try to memorise that fact as a willed goal?), partly about voluntary learning especially in formal education, and partly in between. The in-between is inescapable as many (Piaget, Bruner, ...) have remarked. In the end you cannot explain to a learner exactly why it is worth learning a particular topic: they can only decide that after they have learned it, because knowing what a topic is exactly is to learn it. To a greater or lesser extent this is a large part of our lives; it is why we listen to the news (to learn what we don't know to ask about, but quite often agree afterwards we are glad to have learned), why we have peer-reviewed journals (to raise the chances that it is worth our reading those papers before we know what they are like), and so on.
So motivation can never wholly control learning: important amounts of learning are involuntary. And also, deep learning may always be involuntary in the sense that the motivation was for something else: understanding, not learning. In fact there may be three levels here: a) deep learning, where the motivation is for understanding, b) shallow learning, where the motivation is for learning, and c) amotivation, where the motivation is for some surface behaviour such as doing or looking at what others are doing or looking at.

Costs as well as benefits

The motivation literature seems to treat motivation purely as about rewards or punishments received by the agent as a result of the effects of their actions. In fact in deciding to do something, an agent normally weighs up the payback from the final result against the costs of taking the action. Here, a learner may consider consequences of learning, but also how much effort and other unpleasantness the process (rather than the outcome) may bring. It seems to me a different topic, yet one that is very important to most learners most of the time, to consider costs of process, not just benefits of product.

The issue is in part whether costs and benefits are separately present in the mind, or only a net balance called "motivation". The motivation literature seems to deal only with the latter, while it would seem rational to take the former position. In favour of my argument, to deny the distinction between costs and benefits and to use only the net result would lead to the following. Imagine that you could retire now very rich if only you sold your children into slavery. The motivation literature would describe your failure to do this as "low motivation". This fails to distinguish cases where very large costs and benefits co-exist from cases where both are low (e.g. wanting some chocolate, but not enough to go to the store just for that). On the other hand, self-efficacy and related constructs correspond to the fact that, especially in the case of learning, we humans do seem to experience it as a single psychological variable.

An extreme theory would be: motivation ("do you want to learn?") makes no difference, it is really a side-effect, a synonym for attitudes in the learner ("do you like math?"). But a commonsense view is that motivation causes effort, which causes learning. However we naturally adjust our interest according to perceived reward i.e. we look at the success:effort ratio. What matters about motivation, behaviourally, is effort. (If you learn as a side-effect then effort doesn't matter.)

Behaviour probably depends on the ratio or difference of 2 variables: expected reward, and expected cost (effort, discomfort). Does this cost-benefit analysis make sense, given that a) the effort of attention and thinking seem so small, or at any rate different in kind from external rewards; b) the rewards of being interested i.e. the pleasure in understanding and knowing do not seem comparable to other kinds of reward? Still, in the end cost-benefit analysis seems sensible because behaviour requires a (1-dimensional) choice, so somehow things are compared.

N.B. I fall asleep over TV documentaries when tired, but not when alert. Nevertheless I choose and feel better after viewing them. This implies that the tradeoff does not just depend on externals, but on internal body state. So there are no constants in motivational design! The costs and benefits change from hour to hour, and are not personality characteristics.

Time scales

An important issue that should be discussed at the same time (but often is not), is that of time scale. Some aspects vary, if at all, on the scale of decades. Others (e..g self-efficacy) over episodes. Others can be turned on or off in a second e.g. telling students that this topic will be examined. In fact this latter time scale is more common than we may realise. Once I was going to have to tutor a "lab" using a piece of software, and ran through the worksheet my fellow tutor had written for that week. I went through the first 4 items, then suddenly the fifth demanded that I think about the meaning of the values I had been calculating. With a jerk, I realised that I had not been thinking at all about the meaning, but been fully occupied with finding the commands and following the recipe. The task demands can control the level of processing: I no longer think there is much truth in ideas about "learner characteristics" and deep and shallow learners. These can be changed in a second by the demands of a worksheet. If you ask for a phone number but fail to notice whether it is odd or even, prime or non-prime, does that mean you are a shallow learner? (Great number theorists do notice these things, normal people don't but can.) Many of these things are treated in the literature as if they are long-time-scale features, but this is not usually tested, and in fact is often untrue.

If you administer a questionnaire such as the AMS asking students to fill it in w.r.t. the semester-long course that is one thing; but you could also ask them to fill it in after each little learning activity, and you might find something different. Decisions about action are made all the time, and some (perhaps much) motivation changes just as fast.

Experimental demand characteristics

This is also an issue experimentally. The Hawthorne effect is a standing warning about the difficulties of any investigation of motivation, or of any phenomenon where motivation is a significant causal factor. So, Bev, how do you know in your study what is due to your software and what is due to the effect of a concerned researcher taking a personal interest in the participants?

Missing points not to be lost

There are some important points made in the discussion that I have not absorbed into this synthesis. These are:

1) Joe Beckman's idea that teaching, leadership, and salesmanship are identical in that they depend on "inspiration" which is to do with transferring ownership of an idea from teacher to learner.

2) The overwhelming importance of confidence in language learning. To do with social interaction, and perhaps confidence at succeeding at the surrounding task of communication even with very low technical skills.


Dangerous words

One important aspect that has emerged from this encounter with a difficult area, is that some of the words you meet or naturally use are dangerous: cause you to make conceptual errors. One view of this is that they are ambiguous. Another view is that each such word has a number of independent dimensions. But this is dangerous because we almost always assume, unless explicitly warned and reminded, that one word in English corresponds to exactly one thing. If in fact there is no such thing (e.g. perpetual motion, inteligence, etc.) then we fall into the fallacy of reification: believing there is a real thing there because there is a word for it. If there are several things there (the main problem in this topic), we fall into confusing them.

The words are:


Dweck, C.S. (1986) "Motivational processes affecting learning" American Psychologist vol.41 no.10 pp.1040-1048

Vallerand, R.J. & Bissonnette, R. (1992). Intrinsic, extrinsic and amotivational styles as predictors of behavior: A prospective study. Journal of Personality (60), (3), 599-620.

Vallerand, R.J., Pelletier, L.G., Blais, M.R., Briere, N.M., Senecal, C., & Vallieres, E.F. (1992). The Academic Motivation Scale: A measure of intrinsic, extrinsic, and amotivation in education, Educational and Psychological Measurement, 52, 1003-1017.

Vallerand, R.J., Pelletier, L.G., Blais, M.R., Briere, N.M., Senecal, C., & Vallieres, E.F. (1993). On the assessment of intrinsic, extrinsic, and amotivation in education: Evidence on the concurrent and construct validity of the Academic Motivation Scale. Educational and Psychological Measurement, 53, 159-173.

Bev Taylor's Bibliography on motivation and learning

An old half-cooked model of mine

Keller's References around ARCS

References pertaining to learning motivation:

Keller, J.M. (1979). Motivation and instructional design: A theoretical perspective. Journal of Instructional Development, 2(4), 26-34.

Keller, J.M. (1983). Motivational design of instruction. In C. M. Reigeluth (ed.) Instructional Theories and Models: An Overview of Their Current Status. New York: Lawrence Erlbaum Associates, Publishers, pp.383-434.

Keller, J.M. (1984). The use of the ARCS model of motivation in teacher training. In Shaw, K., & Trott, A.J. (Eds.). Aspects of Educational Technology, Volume XVII. London: Kogan Page, pp. 140-145.

Keller, J.M. (1987). Strategies for stimulating the motivation to learn. Performance & Instruction, 26 (8), 1-7.

Keller, J.M. (1987). The systematic process of motivational design. Performance & Instruction, 26 (9), 1-8.

Keller, J.M. (1987). Development and use of the ARCS model of motivational design. Journal of Instructional Development, 10(3), 2-10.

Keller, J.M., & Kopp, T.W. (1987). Application of the ARCS model to motivational design. In C. M. Reigeluth (Ed.), Instructional Theories in Action: Lessons Illustrating Selected Theories. New York: Lawrence Erlbaum, Publishers, 289-320.

Keller, J.M., & Suzuki, K. (1988). Application of the ARCS model to courseware design. In D. H. Jonassen (Ed.), Instructional Designs for Microcomputer Courseware. New York: Lawrence Erlbaum, Publisher, pp.401-434.

Keller, J.M., & Keller, B.H. (1991). Motivating learners with multimedia instruction. Proceedings of the International Conference on Multi-Media in Education and Training (ICOMMET '91). Tokyo, Japan: The Japanese Association for Educational Technology and the International Society for Technology in Education.

Keller, J.M. (1992) Enhancing the motivation to learn: Origins and applications of the ARCS model. Reports from the Institute of Education (Sendai, Japan: Institute of Education, Tohoku Gakuin University), 11 pp.45-67.

Keller, J.M., & Burkman, E. (1993). Motivation. In M. Fleming (Ed.), Instructional Message Design, 2nd Edition. Englewood Cliffs, NJ: Educational Technology Press.

Astleitner, H., & Keller, J. (1995). A model of motivationally adaptive computer-assisted instruction, Journal of Research on Computing in Education, 27 (3), 270-281.

Keller, J.M. (1996). Motivational design and multimedia: Beyond the novelty effect. Proceedings of the International Symposium on New Technologies of Instruction, 1966. Taipei: Taiwan: National Taiwan Teacher's College. (This document is available in English and Chinese.)

References pertaining to workplace motivation

Keller, J.M. (1992). Motivational Systems. In H. Stolovitch, & E. Keeps (Eds), Handbook of Human Performance Technology. San Francisco: Jossey-Bass Inc., Publishers.

Keller, J.M. (1994). Trends and tactics for employee motivation, HR Horizons, 115 (Winter 94), 5-10.