Working In Uncertainty

Motivation and uncertainty


It's obvious that doubts affect our motivation but you may be surprised at just how powerful the link is, and how it works. This article offers some ideas on how exactly uncertainty and motivation are linked and what we can do to avoid seeing good initiatives fizzle out. (The article doesn't discuss every aspect of motivation; only the link with uncertainty.)

Motivation is something we tend to think about when we're trying to do something that requires repeated effort over a period of time. It's less of an issue with one-off efforts and things that are usually pleasant to do. Nobody worries about being motivated to watch television but most of us have projects at work that people aren't interested in, and personal projects about keeping slim and fit, learning a new skill, or improving our home, for example. With these, motivation is often a big issue.

Have you ever heard someone say "I know I should do it, but I just can't seem to get motivated"? Or how about "I just don't feel motivated today"? Have you ever seen good ideas abandoned, sometimes even after promising early results?

Why do we procrastinate?

One popular theory of why we can feel that we should do something but can't get motivated is that we have a conscious mind that knows what is best, and an unconscious mind that doesn't and needs to be persuaded. The idea is that if we can somehow program ourselves with images, or shape our motives with rewards, then the unconscious will do the right thing.

There may be some truth in this, but there are also some simple, common sense explanations to think about. For a start, we often accept that something would be a good thing to do, but we are unsure of exactly how to go about it, so we dither. We're not sure of what action to take.

At other times the problem may be that in truth we are not as convinced that something is worth doing as we say we are. Often we are right to be unsure. Take exercise programmes for example. In principle being more active through exercise is a good thing for just about everyone. In practice the impact of exercise programmes differs greatly between people, probably for genetic reasons, and the impact of a programme can be incredibly sensitive to details of what it involves. Most men want bigger muscles but the difference between exercise that produces them and exercise that does not hinges on degree of intensity so finely that it is easy to spend months in the gym and achieve almost nothing. So, in principle we accept that exercise should be good for us but in practice we have legitimate doubts about what will be worthwhile.

A simple theory of motivation and uncertainty

Here's the theory, in brief:

Our motivation is affected powerfully and rationally by doubts about whether what we are doing will work / is working, and these change over time in a broadly rational way.

The underlying idea is that our thinking is at least somewhat rational and many of the patterns of wobbling motivation that we are familiar with are the result of expectations and reactions that are, broadly, sensible.

Here's how that works in more detail. Let's assume that there's something with some desirable potential impacts that we would like to try and obtain. However, getting there is going to take persistent effort and involve doing something we haven't done a lot of before. In this situation we will normally have doubts about the effectiveness of our actions and, over time, our motivational problems will change because our needs for evidence of effectiveness change.

  1. Before we act: Before we start taking action all we require are sufficient good reasons for getting started. Perhaps a friend has done it, or a book has convinced us that the course of action is worth trying.

  2. The change that happens when we start to act: However, as soon as we start taking action we begin to expect evidence of at least something happening that is relevant, and we start to look for it. If we can see clear cut evidence that the ultimate results we desire are on their way that is ideal, but failing that we are usually happy to accept at least some evidence of something relevant happening. For example, when trying to lose weight it is encouraging to see oneself slightly lighter on the scales. When you play a computer game there is usually plenty of action on the screen. Just a few button pushes produce fantastic changes and often many points to add to your totals. Psychologists sometimes call this "effectance" and it helps explain why children love computer games, trampolines, and bicycles.

  3. Changing expectations as action continues: As time goes on mere effectance is not enough. We start to expect and want evidence that the ultimate results desired are on their way. For example, when losing weight the scales are encouraging but what we really get a boost from is someone saying "Wow, you look soooo slim." As vain creatures, myself included, being visibly slimmer is usually one of the ultimate results we desire.

  4. Changing calculation as action continues: Another growing worry is whether the results are worth the effort we are putting in. Just getting some results becomes insufficient. Increasingly, we wonder if the results are really worth it.

Sometimes, the connection between our actions and the end results we desire is a cause-effect chain with more than one link. This is typical of long term efforts. Often, it is hard to get evidence of end results at first, but immediate impact is easier to observe. Over time, if we work at it, we can learn to detect impacts further and further along the chain of cause and effect until eventually we reach the ultimate results.

If we fail to obtain at least the minimum required level of evidence of effectiveness at any stage our motivation to continue will begin to drop.

If you think back over the preceding paragraphs of this theory I hope you will agree that all this makes fairly rational sense. Our need for evidence of effectiveness starts low and rises as the opportunity for getting such evidence increases. If we did not think this way it would be a surprise and strangely maladaptive for an animal that has evolved over millions of years.

It also provides an explanation for our feelings about completing a job. For example, suppose you are spending the day putting up a shed in your garden and have not done this before. After unpacking the parts and puzzling over the badly printed instructions you have gradually been making progress - apparently. The walls are up. Part of the roof is in place, just about. Visible progress is there in front of you, but still there are nagging doubts about the ultimate result you desire, i.e. a strongly made shed with all the bits in the right places. If your experience of building badly designed kits with inadequate instructions is anything like mine you are anxious about whether you've perhaps missed a step, or put something on the wrong way around. Until the job is completed and the shed is finished those doubts will probably remain.

If the kit and its instructions were better designed then the visible progress of the shed, coupled with the results of checks suggested in the instructions, would have produced convincing evidence of progress and strong, perhaps even growing, motivation to complete the job.

With these insights it is easy to see how a good idea can falter at some stage even after initial excitement or even promising early indications. Initially we might be persuaded that the idea should work for us because it usually works for other people. However, once we start to act our motivation might change because we discover that the idea does not work for us, or because we cannot establish that it works for us due to lack of evidence, or because we wrongly conclude that it does not work for us based on weak or misleading evidence.

One way that can happen is that, although we can measure changes in the ultimate results, we can't trace those back to our actions, or to specific actions. It may be that there are other factors besides our efforts that are driving results and this clouds the picture. It may be that we are taking a range of different types of action and want to know which of them are really making a difference and which are not. In theory we could be confident that our efforts are producing worthwhile effects but still have doubts about each individual type of action, leading to motivation problems.

How motivation can go wrong

Of course if something really isn't worth doing then we should stop, so just because we stop persisting with something doesn't mean we have made a mistake. However, we do sometimes make mistakes. Here's how:

  • Failing to think through the knock on effects when considering the case for action. For example, we may fail to think through the knock on benefits of something and so underestimate the value of achieving it.

  • Failing to see any perceptible impact from our actions even when they are there. For example, the change may be small or slow making it hard to see without special effort.

  • Initial excitement and motivation dissolving over time because we fail to search for more convincing evidence concerning ultimate results.

  • Faulty measurements of impact leading to a wrong decision to quit.

  • Inability to trace results reliably back to specific actions we are taking.

Avoiding mistaken loss of motivation

With this understanding of how doubt about effectiveness influences our ability to make persistent efforts we can begin to think about ways to avoid killing good initiatives by mistake. I hope you find some useful ideas in this list:

  • Increase continuity of evidence i.e. make sure there is always the required evidence with which to judge effectiveness

    • Increase continuity of evidence over time i.e. avoid periods with no evidence at all

    • Increase appropriateness of evidence i.e. the right type at each phase

      • Initially, good reasons

      • Next, evidence of immediate impact

      • Then, evidence of ultimate results (if not possible earlier)

      • Then, evidence of costs and so cost effectiveness

  • Increase quantity of data obtained

    • Measure more things

    • More often

    • More accurately

    • And summarise less

  • Cut the cost of getting evidence

    • Automate

    • Focus on information already readily available

  • Increase the conclusiveness of data obtained

    • More reliable data

    • Screen out 'noise'

      • Smoothing e.g. with averages or moving averages

      • Accounting for the impact of confounding variables

      • Fitting statistical models

      • By using predicted results instead of measured results when predictions are, in the short term, more reliable

    • Improve experimental design

      • Systematic variation of independent variables

      • Using control groups or otherwise controlling potentially confounding variables

  • Clarify visualisations of progress

    • Make small changes visible graphically

      • Magnify them e.g. with large displays and scales

      • Focus in on the changes rather than the total quantity

      • Chunk up smaller gains into larger units e.g. milestones, blocks, points turning into 'level ups'

      • Colour in boxes with % completion

    • Make slow changes visible graphically

      • Show time series i.e. a graph of how things have changed over a long enough period

      • Animate history, accelerated as with time lapse photography

      • Show cumulative data

  • Look at more types of desired results (giving more chances of finding helpful progress)

Some more practical examples

That was a long list of individual ideas and it may help to think about examples of them used in meaningful combinations to tackle common problems.

Writing an article or report

Writing a document such as an article or report involves developing some ideas and an outline, and also writing the full text. Writing the full text, when the ideas are ready, tends to go quickly and it's great to see the paragraphs building up at such speed. It's easy to see that progress is being made towards the goal of a completed document ready for publication.

By comparison, during the earlier activity of developing ideas and outlines it is much harder to see progress. A mess of confused jottings can get larger without giving a convincing sense that progress is being made.

One way to overcome this is to measure the quantity and organisation of the ideas. Assuming you write your ideas down as lists, track the development of each list through successive versions, including when lists are merged or split. Give a point for each bullet point then extra points for organisation. For example, if a list has 10 bullet points on it but is only 50% well organised in your opinion, give it a score of 10 + 10*0.5 = 15.

Keep a cumulative tally of total bullet points and total score. When a list is superceded by a better version increase the numbers by the incremental improvement. Plot both bullet point counts and scores on a graph.

Calculate your statistics and plot graph points whenever you feel like it.

This procedure turns something abstract, confusing, and difficult to assess into a task with more visible progress. So far so good, but this alone may not be enough.

Suppose the material you have is plentiful and well organised, but not right for the audience. Initial satisfaction with your evidence of progress may begin to ebb away so it is time to start adjusting your scores to reflect appropriateness, perhaps with another multiplier.

Even that may not be enough if it is hard to say what the audience will like and getting it right is extremely important. You may have to get some research and work out a measure of fit between audience interests and your material. The aim as always is to get closer to a continuous flow of improving evidence on whether you are progressing towards the ultimate results you seek.

EVO : evolutionary project management

One of the best ways to deliver any kind of project is with EVO, Tom Gilb's approach to evolutionary project management. EVO involves a combination of rapidly delivering incremental improvements in performance (every one or two weeks for example) and frequently revising the plan to overall completion in light of experience.

EVO is great for motivation. The beneficiaries of the project see some improvement very often, even though it may be small. In addition, Tom emphasises measurement of several forms of improvement, encourages frequent tests of performance, and uses a form of percentage achievement that tends to emphasize the importance of gains made.

Studying for exams

Studying for exams is something you are either doing now or have done in the past. It is one of the severest tests of motivation most people ever face because persistent, usually unpleasant effort is required over a long period, frequently with no convincing sense of progress towards ultimate rewards at all.

Perhaps this is why so many people spend most of the academic year taking notes from which they hope to learn something later, doing the minimum study they can get away with for as long as possible before going with the herd and trying to cram it all in during the last few weeks. What a pity, because this strategy means that the rest of the year is spent going to classes or lectures with only a hazy idea of what has been covered previously and sitting through them slightly confused, taking notes, and waiting for the speaker to stop.

It is far better to be up to date, even slightly ahead of the lessons in your reading and self study, so that each lecture is an opportunity for learning and reflection instead of a dreary note taking chore.

Notes are, of course, a highly visible form of progress which is probably why we focus on them, but the textbooks are usually better written and more accurate.

When I was at school I remember a physics teacher telling us about a pupil he had once taught who took no notes during his classes. According to the story, this pupil was so brilliant that he just read the textbook before the exam and yet still did better than anyone else. At the time I thought it amazing that he had done so well despite not taking notes. Now I can see that he did well in part because he did not take notes. Instead he spent lessons thinking and learning, then for revision relied on the superb textbook. It is quite possible that he also read ahead in the textbook and therefore was well prepared to learn from each lesson.

But how can you motivate yourself, or someone you care about, to do the right thing (whatever that is for you)? Here are some suggestions:

  • Do some research on how exam results will influence the choices you have later in life. For example, when choosing a university or getting a job. Look at your odds in life given several different levels of exam achievement.

  • Experiment with being well prepared for classes. Try going to 5 classes without doing any previous study and 5 similar classes having previously spent 20 minutes looking at a book to preview the material. Keep a note of how much time you spend in every lecture too confused or tired to be learning.

  • Estimate the implications of pre-class preparation for the total time you will spend studying during the year. For example, 20 minutes prep + 60 minutes lecture + 100 minutes revision = 180 minutes, versus 0 minutes prep + 60 minutes lecture (wasted) + 130 minutes revision (no savings this time) = 190 minutes. Tally up the boredom each way. Consider the benefits for your coursework of being up to date all the time.

  • Record and graph simple measures of work done. Work in half hour slots during which interruption is not allowed. Record how many of these you have done. Record how many practice questions you have done. Create a battle plan of the syllabus and mark off how much progress has been made with each part of it.

  • Record and graph measures of growing knowledge. Growing knowledge is more important than growing notes. If you operate a spaced review system this will produce plenty of evidence of material learned. If there is a large bank of multiple choice questions to be mastered then record how many you have done right at least once and how many you have done right three times in a row. Use tests to measure (and strengthen) your knowledge. Measure your speed in recalling key facts and plot speed gains on graphs.

  • Monitor your progress relative to others. In an ideal world your marks in practice tests such as mock exams could be translated into pass/fail or even into grades by a consistent and objective rule. In reality educators find it almost impossible to set exams with a consistent level of difficulty and so it is usually better to look at your marks compared to other people. Discovering how much effort it takes to keep up with your competitors, and how much to start pulling ahead of them, is particularly valuable.

In an ideal world...

In an ideal world, what would evidence look like? Sometimes knowing this helps us design evidence systems for our real world.

Here are some suggestions, illustrated using a fictitious "Homework System", which is a system that watches people studying at home and tells them how they are getting on. It is able to identify improved learning skills as well as increased knowledge.

The system:

  • gathers data without interference to you, the learner;

  • immediately tells you what your ultimate results are across all objectives of interest (e.g. how much more you know, how long the knowledge will last, how much knowledge will remain at exam time, how your learning skills have improved, and all this per unit time);

  • shows how this compares to previous performance and to other people;

  • and at what rate the gains are being made;

  • and what the result would be of putting in various levels of time on the results that would be achieved (e.g. "If you carry on like this but do 2 hrs a week you will be top of your class by July.");

  • and is very clear and easy to understand despite being so informative;

  • and is always perfectly reliable and accurate;

  • able to trace gains back to individual actions;

  • and never gives false feedback driven by unrepresentative bursts of apparent progress.

A real homework system would almost certainly involve some interference to the learner and be applicable to only some study activities. The homework would have to be done at a computer so that every action could be recorded, timed, and analysed. To avoid giving progress indications that were too sensitive to blips in progress it would have to analyse the data statistically and show its results using ranges rather than exact numbers. It would also have to make use of knowledge of typical performance across many similar learners, particularly early on when the data from a learner are still few.


Motivation to continue making an effort with something is greatly affected by our uncertainty about whether what we are doing is working. The type of evidence we expect changes over time. As time passes during which we can acquire convincing evidence of end results we become more demanding of it. This is one reason why motivation can falter even though initial indications have been encouraging.

To be motivated when we should be, and quit when we should, it is a good idea to make sure we are getting the evidence we need, frequently.

Related reading

Expectancy Theory, a similar theory to mine but without the time element, was proposed by Victor Vroom.

"Effectance" is a lovely word and central to Robert W White's theory of motivation. See "Motivation reconsidered: The concept of competence" Psychological Review. 66(5), Sep 1959, 297-333.

Another angle on motivation and uncertainty is covered in "Will-power and uncertainty". The idea here is that the case for continuing with something varies over time and we should anticipate this.

Two slightly more technical articles covering ways to increase learning from experience are "Learning more from experience" and "Better management of large scale financial and business processes using predictive statistics".

For more information on spaced reviews as a way to make good use of study time see "The boring truth about never forgetting" on, another of my websites.

"A new focus for memory improvement" is a guide to memorising that links thinking and memory in practical way.

Tom Gilb's EVO method is described in "Why is Evolutionary Project Management so effective?", which also has links to other excellent sources.

Made in England


Words © 2007 Matthew Leitch. First published 17 May 2007 (slightly expanded 28 May 2007).