242 Quality Assurance
Step 1: Take all measurements in a random order to ensure that any drift or
changes that occur will be spread randomly throughout the study.
Step 2: Estimate all readings to the nearest number possible. At a mini-
mum, all readings should be taken at ½ of the smallest gradation.
Step 3: If operator calibration is suspected to be a major inuence on
variation, the gauge should be recalibrated by the operator
before each trial.
Step 4: The study should be observed by someone who understands the
importance of the precautions required to conduct a reliable study.
Trials and Data Collection
The actual process begins by making sure that
• There are generally two to three operators.
• There are generally ve to ten process outputs to measure.
• Each process output is measured two to three times (replicated) by
each operator.
To use the variable data MSA, one must begin with the worksheet of the soft-
ware (Figure 16.6). Here we enter the raw data. Once the data is entered, then
we follow the rubrics for analysis in the manner shown in Figures 16.7 and 16.8.
Manipulate the Data
Once the data is in the work sheet, now we go to the commends to stack the
data. In Minitab
®
, this is a very important step. If stacking is not selected, the
analysis may not be correct.
(randomized order)
ocess outputs
erators
Have Operator 1 measure all
samples once (as shown in the
outlined block)
en, have Operator 2 measure all
samples once
Continue until all operators have
measured samples once (this is
Replicate 1)
Repeat these steps for the required
number of replicates (parts in
random order)
Enter data into Minitab
®
columns as
•
•
•
•
•
FIGURE 16.6
Data in worksheet.