Matthew W Ford

Research

Service

Home

Deming's Bead & Funnel Demos (Updated 01/21/2009 09:13 PM)

We've identified W. Edwards Deming as one of the "gurus" of quality management.  We also noted that one of Deming's contributions to the field of quality management was the concept of understanding variation (a topic that we're currently engaging in our discussion of quality control).  To teach others about variation, Deming used to conduct two demonstration experiments with his students.

The Red Bead Experiment.  From an audience of onlookers, Deming would select several 'willing workers,' a recorder, and a couple of inspectors (Deming, 1986).  The workers were instructed by the foreman (usually Deming) to dip a paddle with 50 holes in it into a tub of small wooden beads (800 red/3200 white).  The results from each worker's dip would constitute one day's production.  The performance standard was to draw 50 white beads from the tub when the paddle was dipped.  The foreman also noted that each workers' job depended on their ability to produce only white beads.

The workers take turns drawing beads from the tub.  The inspectors count the number of red and white beads on the paddle.  The recorder writes down the numbers and plots the values on graphs drawn on a white board or flip chart pad.

The foreman evaluates the results.  As production progresses, the foreman tries the following things to improve worker productivity: berating workers who draw higher numbers of red bead for poor performance, implementing a Zero Defects Day to encourage better worker performance, hanging posters that emphasize the importance of quality, and firing workers who consistently miss their work quota.

The Funnel Experiment.  A funnel is suspended above a table.  Below is a tablecloth with a target drawn it.  Participants drop marbles through the funnel, with the goal of hitting the 'bullseye' in the middle of the target.  After each drop, the distance of the marble from the bullseye is recorded.

Four different 'dropping' strategies are examined:

1. Leave the funnel alone and merely drop marbles thru the stationary funnel.
2. Measure a dropped marble's distance from the bullseye and then adjust the funnel in an equal but directly opposite direction.  Repeat after each drop.
3. Measure a dropped marble's distance from the bullseye, center the funnel over the target, and then adjust the funnel in the opposite direction of the error from the target.  Repeat after each drop.
4. Place the funnel over the spot where the marble lands.  Repeat after each drop.

Partial results reported by Deming (1993) appear below:

Graphs of typical results from these four rules can be viewed here.

Answer the following questions in a one page memo:

1. What are the red beads supposed to represent in the experiment?
2. What type of variation (common or special cause) do we observe when we draw different numbers of beads from the container each time?
3. Over the long run, what percentage of red beads should we expect from this process?  Explain.
4. Will each paddle draw the same percentage of red beads?  Explain.
5. Will long term results from the red bead process change as management continues to urge the workers to work harder and to produce less mistakes?  Why or why not?
6. If managers want to reduce the percentage of red beads over the long term, what do they need to do?
7. The key lessons related to the funnel experiment relates to 'tampering.'  What do you think tampering is and how does the funnel experiment relate to tampering?  Which of the four strategies above involve the least amount of tampering?
8. What is a control chart and how might it help managers avoid tampering?

References

Deming, W.E. (1986). Out of the crisis. Cambridge, MA: MIT Center for Advanced Engineering Study.

Deming, W.E. (1993). The new economics. Cambridge, MA: MIT Center for Advanced Engineering Study.

Site maintained 1999-2013 by Matthew W. Ford.