Writing a SPC report for managment

What are you struggling with? Perhaps if you post an example?

I also recommend Donald wheeler's on www.qualitydigest.com

 

There are a lot of hack teachers of SPC (and other quality methods) out there so I wouldn't worry about your undertanding yet. I've seen a lot of students provide summarization of an SPC case study that got As from their professor yet their answers were total crap from a real life perspective. I look forward to seeig your example...

 

First, go beyond stating the 'statistical' obvious and describe the condition in more detail. for example, it is clear that there is a bias to the reference value. but the unanswered questions are: what is the direction of the bias? in this case the smaller the value, the system measures high to the reference value and the higher the reference value the system measures the parts low to the reference value. Is this bias of any practical importance? the bias is about .5 at it's worst. what are the spec limits? does this bias matter?
The second thing to think about is: what is graph about? and what do the results mean. what is the practical effect? for example, the Range chart shows the range of measurements (max-min) for each part and operator. it is a direct display of repeatability.

now you try it with the Range chart for operators. what do you see?

 

Bev's second thing to think about is more meaningful for management, in my opinion. You're taking an SPC course and learning to understand the story that the numbers are telling. Management, however, doesn't necessarily speak fluent "stats" and your role may be to translate the story so that they understand it and possibly recommend actions that could get results back to where they should - or need to - be.

Repeatability is lovely...but are the results repeating in a way that management wants?

Are you trending up or down? If you stay on this pace, when will you hit the desired target? Or worse, when will you fall outside of acceptable range?

The numbers can tell a story. That's what the report to management should focus on, in my opinion.

 

It can be difficult if you don't understand (or are not given) the application of the numbers (i.e., the manufacturing/service process it represents). That's the downside to classes that focus on "application of the theoretical"...there's a need to balance it with the "application of reality."

Can I presume that the typed content on page 2 of your attachment is your analysis? If so, let's pretend I'm management.

• I've read this and I'm thinking "So what" - how does it impact me as management?
• Evidence of bias...no evidence of bias. What does this mean with regard to the instrument? Do I need to buy a new one?
• "Operators measuring system might need to be looked at" - For what? Why? What is wrong with the current system? Wouldn't training work?
• What does non-linearity mean in terms of a process? If my equipment or gage R&R system is flawed, how does that impact my manufacturing process or, more important, my final product that the customer wants tomorrow?
• I don't understand why different gauges should be used because of a line's slope.

Hopefully, these types of questions will help you translate from the language of numbers to the language of management. Good luck!