 |
 |
Online Quality Assurance and Manufacturing Data Management |
|
|
||||
|
|||||
|
Mean: This statistic is simply the average of the samples. The mean measures the central tendency of the data. Note that basing a decision on simply looking at a mean value without considering other indicators can lead to erroneous conclusions. Standard Deviation: A measure of variability of the samples set. A large standard deviation indicates a sample set with much variation and a small standard deviation would indicate a consistent sample set with only small variations from the mean. Minimum, Maximum, and Range: The minimum and maximum values measure the extreme values in a sample set. The range is simply the difference between the maximum and minimum. It is of interest to compare the manufacturing specification range to the measured range. Note that the range by itself is not a good measure of variability. Count: This statistic represents the total number of samples. It is important to note the sample size to assure that sufficient samples have been evaluated to make a statistically valid interpretation of the data. 3 & 4 Sigma Limits: The mean +/- 3 times the standard deviation represents the 3 sigma limits. Assuming a normal distribution of the data, there is a 99.74% probability that all of the samples will fall within the +/- 3 sigma range. Similarly, at +/- 4 sigma, there is a 99.9% probability that all of the samples will fall into this range. % Above and Below the Specification Level: These numbers simply indicate the % of all samples measured above the upper specification level (USL) or below the lower specification level (LSL). CPK: This statistic is a measure of how close the process is running to the limits of the specification range. The statistic is influenced by both the mean and the standard deviation of the sample set. A CPK value near one or above is desirable since this indicates the process is running with the manufacturing specifications.
|