Use the capability histogram to assess the normality of the data and to compare the distributions assuming the within-subgroup variation and overall variation. The histogram of measurements consists of the following:

·    Within and overall curves that are superimposed on the histogram. The within and overall curves are normal density functions using the process mean and different variances. The within curve uses the within-subgroup variance, while the overall curve uses the overall sample variance.

Assess the normality of the data by comparing the curve to the bars. A normal distribution is symmetric and bell-shaped. With small sample sizes, it can be difficult to assess normality with a histogram. The piston data appear approximately normal.

Examine the curves to see how closely they follow each other. A substantial difference between the within and overall curves may indicate that the process is out of control, or that there are sources of variation not estimated by the within component. For the piston data, the within and overall curves are closely aligned.

·    Target, and lower and upper specification limits represented by vertical dashed lines on the histogram. Compare the histogram bars to the lines to assess:

-    Whether the measurements are centered on target

-    Whether the measurements are inside the specification limits

For the piston data, the process is approximately centered on target and the measurements are located within the specification interval.