Tuesday, April 25, 2006

The truth about the Red Bead Experiment, as I see it

red beadswhite beadsIn a previous post, I mentioned W. Edwards Deming and how most of us ignore the lessons he taught the world about quality. A friend of mine wrote me back and talked wistfully about Deming's famous Red Bead Experiment -- how that illustrates the ignorance of quality shared by management in much of American business, industry, and (as I now know very well) public school systems.

I used to teach the Red Bead Experiment in my TCNJ classes (and maybe I should bring it back -- I still have the beads, after all), though strictly speaking, what I did wasn't exactly what was outlined by Deming according to most sources (like the Web site I link to above). Here's what I'd do:
  • I would have a small bucket of beads, and about ten percent of them would be red. The rest would be white. (This far I'm with Deming, except if I read him correctly he used red beads that were also a slightly different size from the white ones. I had to get same-size beads at the local craft supply store.)
  • I'd choose five students to be "workers," one to be "management."
  • Each in turn would use a 1/2 cup measuring scoop to get a level scoop of beads. (I usually ask the students to reach in without looking. Lost a fair number of dropped beads that way. Very realistic. But sometimes I would let each player look. I found that didn't change the results significantly. I different winner in a given round might well still be the loser in the next.)
  • We'd count the number of red beads each got.
  • The one who got the most was "fired" and the one who got the fewest got a "raise."
  • Return the beads to the bucket, and shake up the bucket.
  • Repeat the process with the scoop. Inevitably the one who got the raise would be just as likely to be fired in the next round as to get a raise again.
This is simpler than the experiment the Web sites describe; in my classes, since they're not about statistics or quality, it's not necessary for me to get into control charting and upper and lower control limits. But I don't usually need to get all the students fired before the rest of the class gets the idea: many of the "standards" we use to evaluate employees and co-workers are outside of their control. (Young people use the word "random" very loosely these days, but it applies here.) And some of the things they can't control, WE can help them with, through a combination of extra (and usually simple) training, mentoring, etc.

What takes a little extra effort for us to teach is this: management is responsible for anticipating what employees won't know, what questions employees might ask, what system problems might cause employees to "under-perform." You can't leave that in the hands of employees, because they might not know enough about the system to know what questions to ASK about the system. I gotta tell ya, if we leave employees to figure out for themselves what they come in to the job not knowing, we're asking them to avoid red beads. It's cruel.

Labels: ,

posted by Ron @ 7:38 PM

0 Comments:

Post a Comment

<< Home