文档介绍：Are you meeting your target?
What is your reject rate?
Is it 10% less than last year?
What is the availability of the new machine?
Is it equal to 95%, as promised by the supplier?
What is the inventory level?
Has it decreased since we reduced the cycle time?
Using data to assess the level of performance
cannot be totally accurate, because of the
natural variation of the process.
What is the average cycle time?
Is it less than the target of 10 days?
Assessing the reject rate
The reject rate for a process in the first six weeks of
1999 was:
Week 1 2 3 4 5 6
Reject rate % % % % % %
Can we use this data to estimate the average
reject rate for the whole year?
Can we be confident that the reject rate is less
than 10%?
Mean or average = % Median = %
Standard deviation = % Range = %
% % % % % Reject
rate
The data
If we simply predict that the reject rate is %, then:
· Our prediction will be wrong.
· Furthermore, other people cannot assess how
wrong our prediction is likely to be.
Measuring variation
Range = Largest - Smallest
% % % % % Reject
rate
Range = %
The range is the width of the dotplot
Measuring variation
Standard deviation is an alternative measure of
variation.
It is better than the range because:
· Larger samples tend to have larger ranges.
· The range is very influenced by the smallest and
the largest values.
Standard deviation = Ö[S(x - m)2/n]
or
Standard deviation = Ö[S(x - m)2/(n-1)]
The standard deviation is better
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
Mean =
Range =
SD =
Mean =
Range =
SD =
Mean =
Range =
SD =
The long term reject rate
The reject rate for the six week period was %
We need to estimate the long term reject rate.
Confidence limits for the long term rate are given by:
m + ts/Ön
Where m is the sample mean
s is the sample standard deviation
n is the sample size
t is obtained from Table ST1.