But don’t obsess over the distribution.

Yes, I said it. Normal data are nearly never normal.

In Six Sigma classes we study outliers, shift, drift and special cause events. But what we don’t always consider is that these “unexpected” data points may be part of the process and not as rare as we think.

First, let’s look at a set of screw torque data. The chart in FIGURE 1 is for a set of screw torques taken sequentially from a “smart” driver. We can see the data are normal (p=0.895), and the histogram and time series plot back that up.

One measures the variability of process outputs. The other assesses the tests themselves.

People often confuse measurement system analyses and capability studies. Far too often, I hear, “When will we run the capability study on the tester?” And while I’m sure those few brave souls who read my column do not fall into this trap, you might know of people who do. Maybe this column will help.

MSAs are for tests and gages. Capability studies are for the processes being measured. Or, to state it another way, MSAs give us confidence we can measure the capability of our process to produce parts to our customer’s specification.

One can talk about the “capability” of a tester, but only when the word is being used in its classic sense,  for instance, “the extent of someone’s or something’s ability.” Let’s review.