Disconcerting Unease with Math and Statistics


    Management in Western Countries seem to have a very disconcerting sense of unease over Math and Statistics. There is one simple and best way to handle it.

    Get over your fears and learn some more of it!

    Edward Deming is famous for one of the experiments he always ran first in his seminars on quality.

    There was a bowl in which there were 800 white and 200 red beads. They were all mixed together and a cup is provided to measure out approximately to a blue line in the cup. You expect 50 marbles. Few get it exactly right since the measuring cup is very imprecise. Some get 50, some get 49, some 53 and so on.

    The point made is that the problem is with the system (imprecise measuring device) rather than the desire of the worker to perform well.

    Someone who gets 65 may be filling it way past the Blue line. Someone getting in 40 may be underfilling it.

    The idea here is that as long as the random variation inherent within the system is not over the Upper Control Limit (say 55) or below the lower control limit you set (say 45), you let the situation pass. If they go way over it or way under it you make some changes or instruct the workers properly. Statistics may help you define these Upper Control Limits and Lower Control Limits in very precise terms but you get the idea behind it.

    There is also another experiment, (the funnel experiment) that proves that you can make situations worse by making fixes for every expected minor variation you see rather than the ones that do matter.

    All of which goes to prove that just lurking behind the math and statistics of it is some common sense that is very useful in running stable systems – manufacturing or services!

    They are not just some simple idle geek talk! If you care to stabilize your systems and improve your yield and quality over the long run. it is just as important to understand the Zen of these methods as much as the Math and Stats of it!

    However, if you want to produce a car that lasts 20+ years and 200,000 miles your Math and Statistics better get very very precise since that’s how you ensure the parts that make up a car fit perfectly together with extremely low tolerances, increasing the overall quality by making individual quality levels high.

    Same principles apply to outsourcing, especially Business Processes. You better have very precise expecations that can be measured and monitored so that the customer experiences are consistent and meet the expected quality level with variation reduced to a minimum. Variation costs money!

    Pure mathematics is, in its way, the poetry of logical ideas. ~Albert Einstein




    Please enter your comment!
    Please enter your name here