Management is Prediction

Posted on July 14, 2005  Comments (10)

re: post on prediction [link broken, so removed] on the Deming Electronic Network:

Petter Ogland wrote:

…that intelligence more or less boils down to updating a predictive model of the world. As far as I can see, this is the C.I. Lewis epistemology that Shewhart and Deming based their philosophy upon.

…but is there any kind of operational definition for ‘prediction’ that would explain what Deming means when he uses this word in various contexts?

I think your first point is correct, which I see as: learning by predicting, then looking at the result and then adjusting understanding to this new information is very powerful.

I believe Deming’s thoughts about prediction are most effectively put into action using the PDSA cycle. Specifically, you must predict the results in the planning phase (prior to piloting improvements). I find that this is rarely done. I don’t think the form of that prediction is critical (narrative with loose numerical guesses, precise numerical prediction…). The critical issue is making the prediction, then comparing the results to that prediction and then figuring out how your original understanding can be improved based on the new data.

Learning will not only be about the specific case being examined, but also, over time, learning about your tendencies in prediction. For example, do you: overestimate the size of the improvement, underestimate the time it will take to institute a new process improvement, underestimate the complexity of IT projects…? Basically, over time, learning from your prediction history the biases that affect your predictions (which are then tied to your model for viewing the world…). Then, that knowledge can be used to improve your ability to make better predictions in the future. Without actually predicting and then examining the results people often make the same errors in their belief about the potential outcomes of changes for not just years, but decades.

I believe the act of formally making a prediction is critical to improving the learning process. I think prediction and examination of results is rarely done. And I think it is a very powerful component to creating an organization that can improve rapidly.

The understanding Deming had of prediction was as a component of the Theory of Knowledge. The importance of prediction is due to the impact it has our our learning. Prediction, with subsequent evaluation of results and then adjust to the model used to make the original prediction is a process that improves a person’s ability to understand how they think. By going through this prediction cycle many times the person becomes better at using the knowledge the have. Even if they don’t learn about the theory of knowledge abstractly as a concept they learn about the theory of knowledge in practice. On page 101 of the New Economic and for several pages thereafter Deming addresses prediction.

John Hunter

10 Responses to “Management is Prediction”

  1. Curious Cat Management Improvement Blog » Blog Archive » Six Keys to Building New Markets by Unleashing Disruptive Innovation
    July 4th, 2006 @ 11:14 am

    […] Very well put in my opinion. Unfortunately we do not often stop to examine the theory and therefore fail to learn as we should as we experiment and get new results. The most effective way I know of to improve the learning is to use the PDSA cycle, predict the effects of change and then examining the results of change. As simple as those two acts are, they are skipped far to frequently. […]

  2. Curious Cat Management Improvement Blog » The Illusion of Understanding
    November 25th, 2006 @ 6:04 pm

    It is important to understand the systemic weaknesses in how we think in order to improve our thought process. We must question (more often than we believe we need to) especially when looking to improve on how things are done…

  3. Knowledge Management - Management is Prediction
    January 21st, 2007 @ 9:33 am

    […] The most important factor to me, is the value to improving management. The technology discussed here eventually could help lead to the adoption of Dr. Ackoff’s ideas on capturing corporate decision making which will allow evaluating decisions for systemic weakness and strengths to improve decision making. […]

  4. CuriousCat: Predicting Improves Learning
    May 10th, 2007 @ 7:07 pm

    Personally I have found the act of actually making predictions and examining the results incredibly helpful in improving the speed and depth of my learning…

  5. CuriousCat: How to Get Ahead
    June 27th, 2007 @ 2:11 pm

    I find that who says something is most often more important in predicting how people will react than what is said…

  6. Cliff Norman
    May 28th, 2011 @ 11:47 am

    Great discussion. The prerequisite for a prediction is some theory in use. In using the PDSA cycle, is important to frame questions that help people explicitly state their theory, especially if a team in involved. Merely answering yes or no, or guessing, does not make the theory explicit. I have seen instances where people will predict “yes” or “no” for two different reasons. The act of making the prediction can change how we collect data or run a test. Richard Feynman commented, “Science begins and ends in questions.” So should the science of improvement. Questions lead to predictions (theory being explicit), this leads to a plan for data collection and a test. Built into the PDSA cycle is the logic of deductive and inductive learning; Plan to Do (deductive)Study to Act (inductive). This iterative learning process was explored in Statistics for Experimenters by Box,Hunter and Hunter.


    Cliff Norman (API)

  7. lepine kong
    January 1st, 2012 @ 12:22 pm

    One should remember that Deming and Shewhart always refer implicitely to probability theory from which they created spc (statistical process control).

    The one main idea of spc is that you cannot make prediction like you learn at school with normal law that 95% of datas will be between 2 standard deviations if your process is not under control.

    To achieve this state control and so be able to make prediction, they derived another tool PDSA which is a layman term for induction / deduction scientific process inherited since at least Aristotle circle.

  8. From Lean Tools to Lean Management » Curious Cat Management Improvement Blog
    September 3rd, 2013 @ 9:53 am

    Jim Womack: “Only management by science through constant experimentation to answer questions can produce sustainable improvements in value streams…”

  9. Deming Institute Conference: Tom Nolan » Curious Cat Management Improvement Blog
    October 24th, 2013 @ 7:34 pm

    […] also discussed the 100,000 lives campaign. Quotes he used: “Management is prediction” – W. Edwards Deming, “There is no substitute for knowledge” W. Edwards […]

  10. Deming 101: Theory of Knowledge and the PDSA Improvement and Learning Cycle « The W. Edwards Deming Institute Blog
    December 5th, 2013 @ 12:07 am

    Ian Bradbury: “The really critical piece of the [PDSA] cycle then is like we had in the story of Chantecler – prediction: prediction and the implicit theory upon which the prediction is based. So you are making change that is aimed at improvement but you are doing that at the same time as making predictions about outcomes articulating the theory. And then when outcomes are not as predicted… then we have cause to change the theory – cause to learn.”

Leave a Reply

  • Recent Trackbacks

  • Comments