Attribute MSA Classroom Example Ideas

I was just discussing attribute MSA today with an MBB and I was searching for an example of a good demonstration. What about bringing in a bunch of used golf balls and asking your "inspectors" to judge the "play worthiness" of the golf balls. You can give them some general criteria about what "play worthiness" means, but leave it vague enough that people have to interpret the rules. Number the balls with a marker (you'll need 30 or so for the test) and then you should pre-determine which are good and which are bad. You should have about a 50/50 mix.

Of course, it helps if you play golf and have a collection of golf balls. You can use criteria like how clean the ball is, how old the ball looks, how many scuffs the ball has, even what brand, maybe something silly like "no odd numbers on the ball" (not including your reference number). As a golfer, my own criteria is pretty picky, that's why I have such a large collection of "perfectly good golf balls" (according to my wife, who doesn't play).

Have each inspector randomly pull each ball from a bag and inspect it and check pass or fail on a sheet. Have them do it quickly, like 5-10 seconds per ball, or better yet have them sit at the same table and you pull each ball out and show them all at the same time. That way you control how long they see each ball.  Do the test twice, each time with a clean sheet so they aren't influenced by their prior result. You can calculate the results with a spreadsheet or Minitab.

If you decide to use this one, let me know how it works out.

You can do the same exercise using M&M's.  Write an operational definition for different types of defects, like chips in the candy coating, legibility of the "M" printed on each candy, flat spots or mis-shaped candies, etc.  Then perform the measurement system analysis and revise your operational definitions accordingly. 

I thought about M&M's. The problem with them is how do you identify them so that they can be randomly picked but that you can still identify it's known state. I suppose you could put each one in an envelope and number the envelopes to keep track. The only other concern then would be good M&M's being damaged during the inspection process. That might serve as a lesson too though. This whole thought process would make for good classroom discussion, wouldn't it? 

what would be the expected outcome of this?

Your last 2 sentences sum it up.  We teach our GBs and BBs with the M&M excersize.  We just tell them that they have to figure out how to identify the samples and keep track of the 'truth state'.  It makes for great debriefs after the excersize and people make lots of mistakes that they won't make again in real life.  We also don't tell them what defects to look for.  We tell them that they have to define it all themselves.

Troy:

I recently saw a classroom example for MSA using cups of water (3 cups of water, three 'measurers', repeated twice).  An xbar and r chart was created from the data for the class debrief.  It seemed like a great, simple example, if someone who has seen this and has a grasp of the concepts could outline it for you.

We use M&M's to run a GR&R for our GB class.  The way we keep them straight is to use small paper cups with a number (1-30) on the bottom.  This also ensures that the repeated handling of the M&Ms does not degrade the logo. 

We then ask each of the inspectors to group the M&M's into an Accept and Reject group based on their interpretation.  The reject group is then passed to the data recorder who checks the numbers on the bottom of the cups and logs accordingly.  The cups are then randomized on the table and the next inspector proceeds.

Hope this helps!

Bob J

Troy,

This one is easy....what does your organization do?  What do you check or verify?  For example, if there are functions within your organization that actually inspect, check, evaluate, etc., then set up the check using a teaching example for the class.  Another example, you work in the mail room and you must put letters in the correct mail boxes.  Get a sample of ten to twenty letters with names & addresses (some will have defects).  Show them to the group based on some acceptance criteria, then have them assess the samples.  Record the outcomes.  Pick three or four students to input their results in a Excel spreadsheet and look for agreements and disagreements.  Follow the traditional short attribute R&R study outlined in the AIAG MSA book or the one that Minitab uses.  Walk them through the analysis and your are done.  Ask the appropriate probing questions about sample size and selection.  What would be an acceptable standard?  How could they use this tool on their projects, etc.......I think you get the idea.  Good luck!

Troy,

We use a mixed bag of six sigma event charters from previously run activities.  For new students, it gives them the ability to evaluate multiple charters (that they will ultimately held accountable to produce themselves) plus it gives us instructors the ability to introduce our charter evaluation criteria after the students strive to extemporaneously draft their own.  Works great!!

Larry:

The water measurement exercise is a very good one - applicable across practically any industry, but ours is an objective MSA example where the level is objectively measured with differing gauges of varying quality and discrimination.  Is there a subjective study out there for water measurement that I'm not aware of?

We use MM's for our GB and YB classes as well, except that we have the students decide if the MM''s are either "shiny" or "dull", dull mm's being the defect in question.  Everyone always seems to have different ideas about what is shiny or dull, so it generates a lot of discussion about the problems with attribute data and why MSA is so important.

I'm getting a little bored of the MM's myself as an instructor though....(I've taught too many classes I guess) I came here looking for a fun new example!

Nick

what is M&M example.may persons are using this as seems in discussion fourm.Pl explain me or  send detail of how it worked.Anyone pl.

what is M&M example.may persons are using this as seems in discussion fourm.Pl explain me or  send detail of how it worked.Anyone pl.

I use an exercise where the class are coin inspectors for the mint.  Their job is to determine on one side of the coins they are given, what the op def's are for non-conformances.  Then the team is given a set of 20 coins of which has both "conforming" and "non-conforming" coins.  They select two team members to be "master" inspectors and determined the "true" value for each coin.  Then three other members have to rate each coin twice in different random orders.  We make an attempt to keep the operators blind as well.

This exercise works great unless you forget your coins.  Then I default to the M&M method since finding M&M's while on the road is pretty easy to do.