posted 16 October 2000 02:27 PM              

---

I have downloaded the R&R spreadsheet on this site. I am familiar with some analysis of variance software but I'd like some guidance with this spreadsheet. For one particular lab test, I have four machines that should output similar results. I have 8 lab techs. I planned on each tech running 10 standardized samples through two machines with each machine being a "trial". Is this the proper way to address this? What variabilities will the outputs represent if I do it this way?

IP: Logged

Posts: 4119
From:West Chester, OH, USA
Registered:

posted 18 October 2000 06:17 PM               

---

I guess we should start out by asking which R&R spreadsheet? What's the file name?

If all the machines are the same brand, etc., you *may* get away with doing an R&R on 2 of the 4. But - are you doing an R&R on the machine or on the instrument measuring the output?

IP: Logged

posted 19 October 2000 03:22 PM              

---

The spreadsheet I downloaded is RandR.XLS. It appears to be very similar to gagr+r.xls. In my terminology, machine = test equipment. The manufactured "part" is beads of plastic resin. I used a standardized material tested in two of the machines. The material is very uniform and can be considered to have minimal variation.I will likely increase the study to all four machines.

IP: Logged

Posts: 4119
From:West Chester, OH, USA
Registered:

posted 19 October 2000 07:47 PM               

---

I have seen this type of 'similarity' argument successfully used - but I would think with only 4 devices I would do an R&R on each.

IP: Logged

All times are Eastern Standard Time (USA)