Our World Statistics Day conversations have been a great reminder of how much statistics can inform our lives. Do you have an example of how statistics has made a difference in your life? Share your story with the Community!
Choose Language Hide Translation Bar
Highlighted
Level II

## Measurement System Analysis Approach

Hi together,

I`m struggling with the design of a gauge R&R analysis. Maybe some of you could advise me on this.

When my process produces different products with different specifications and all products are controlled by the same gauge.

Does it make sense or is it common to conduct one gauge R&R analysis including the different parts from the different products?

Or is it common to test the reliability of the gauge considering only parts from the same product?

2 ACCEPTED SOLUTIONS

Accepted Solutions
Highlighted
Level VII

## Re: Measurement System Analysis Approach

I assume you mean all products are measured by the same gage?  There are a number of options to conduct a gage study:

1. Take multiple parts and measure the same dimension with the same gage multiple times (often with multiple inspectors),

2. Take multiple dimensions from multiple parts and normalize the difference in the dimensions(e.g., use the delta Target-actual)

3. But, what questions do you have about the measurement system? What conclusions do you wish to draw from the study and over what inference space do you wish to draw those conclusions?

So for example, you question whether operator technique biases the measurement, you would need to include mutiple inspectors in the study, If you question whether environmental factors effect gage variability, you would need to study the gage over changing ambient conditions.  If you would like to be able to determine if the gage can assess within part variation, you would need multiple measure within part, if part-to-part variation is of concern, then multiple patrts, if lot-to-lot variation is of interest, the parts must come from multiple lots.

I suggest you get Wheeler and Lyday's Evaluating the Measurement Process for further information. Referenmced in the JMP White Paper https://www.jmp.com/en_in/whitepapers/jmp/emp-management-systems-analysis.html

Highlighted
Level II

## Re: Measurement System Analysis Approach

Thanky you very much. The webinar helped a lot!

3 REPLIES 3
Highlighted
Level VI

## Re: Measurement System Analysis Approach

The experimental design can be created using either framework. If you use parts from only one product, then your design would have not tell you anything about the other products. You can incorporate the different products and if you create the products as a level in the design, then you'll have information about parts, nested within the products. Other considerations are typically operators, crossing parts/products across the design etc. You may want to take a quick look at this on demand webinar I created during my tenure in the JMP division of SAS.

https://www.jmp.com/en_us/events/ondemand/mastering-jmp/evaluating-and-monitoring-your-process-using...

Highlighted
Level II

## Re: Measurement System Analysis Approach

Thanky you very much. The webinar helped a lot!

Highlighted
Level VII

## Re: Measurement System Analysis Approach

I assume you mean all products are measured by the same gage?  There are a number of options to conduct a gage study:

1. Take multiple parts and measure the same dimension with the same gage multiple times (often with multiple inspectors),

2. Take multiple dimensions from multiple parts and normalize the difference in the dimensions(e.g., use the delta Target-actual)

3. But, what questions do you have about the measurement system? What conclusions do you wish to draw from the study and over what inference space do you wish to draw those conclusions?

So for example, you question whether operator technique biases the measurement, you would need to include mutiple inspectors in the study, If you question whether environmental factors effect gage variability, you would need to study the gage over changing ambient conditions.  If you would like to be able to determine if the gage can assess within part variation, you would need multiple measure within part, if part-to-part variation is of concern, then multiple patrts, if lot-to-lot variation is of interest, the parts must come from multiple lots.

I suggest you get Wheeler and Lyday's Evaluating the Measurement Process for further information. Referenmced in the JMP White Paper https://www.jmp.com/en_in/whitepapers/jmp/emp-management-systems-analysis.html

Article Labels

There are no labels assigned to this post.