Page 130 - Six Sigma Demystified

Chapter 5 m e a s u r e s tag e 111

ments were established to eliminate potential error between personnel, so no
R&R analysis was performed on the cycle time metric.
Likewise, each error type was clearly defined to avoid discrepancies. Yet the
team acknowledged that the baseline data conceivably could be compromised
by inaccurate count of error owing to simple miscounts or an error in interpreting
the correct license count, for example. The team did a quick study using two team
members to review 50 orders. Each came to the same conclusions, with no ob-
served differences in estimates.

recommended Tools

The following tools (discussed in detail in Part 3) are applicable to the measure
stage of DMAIC:

Process Definition

• Flowcharts and process maps to define the process- evel activities necessary
l
Process Baseline Estimates

• SPC control charts, including C, Np, P, U, individual- X, X- bar, and EWMA
charts, for investigating process stability and evaluating process capability
• Histograms to graphically display the process output relative to the
requirements
• Confidence intervals on mean and confidence intervals on proportion for es-
timating process performance when the process is not in statistical control

• Goodness- of- it tests to verify the statistical distributions assumed in the
f
various statistical tools

Measurement Systems Analysis

• R&R studies to quantify the measurement error associated with the equip-
ment, personnel, and procedures
• Linearity analysis and regression analysis to understand the measurement
system error as a function of measurement size

In addition, the consensus- building tools noted at the end of Chapter 4 are also
applied as needed.