1. Hanford Performance Indicator Forum
Life Cycle of a Trend
Hanford Performance Indicator Forum
October 31, 2007
Steven S Prevette
Fluor Hanford ESH&Q

2. Establishing baselines What to do following a Trend
Life Cycle of a Trend
Establishing baselines
Detecting a Trend
What to do following a Trend
How to re-baseline
Setting Goals

3. Introduction This presentation provides detailed information on
Establishing baselines
Detecting a Trend
What to do following a Trend
How to rebaseline
Hanford PI Forum

4. Trending Cycle Take action to correct, reinforce, or stabilize
0. Gestation:
Choose a Measure
4. Maturity:
Data stabilizes
Take action to correct, reinforce, or stabilize
Establish Expectations,
Goals, Routine Monitoring
1. Establish Baseline
3. Youth:
Determine special cause(s)
2. Birth:
A trend is detected
Hanford PI Forum

5. Similarity to PDSA Cycle
Act on large scale or abandon
Plan a change or test
Stable Baseline,
Review Pareto, Histograms
Stabilize on improved baseline
PDSA from The New Economics, Deming
Detect Improving Trend
Study results, what did we learn
Do – carry out on small scale
Hanford PI Forum

6. Gestation Decide what to measure Develop Operational Definition
Develop Operational Definition
Independent data, and do not cumulate
Determine data source(s)
Gather initial data, hopefully at least 25 data points of history
Hanford PI Forum

7. Gathering Data Check existing reporting systems.
Many organizations are awash in data, but the data have never been analyzed or put to use.
There is an advantage to using existing data, as it has already been paid for, and historic data for establishing the baseline should be available.
Hanford PI Forum

8. Choosing Reporting Intervals
If a trend developed, how long could you go without needing to know it? Longer intervals imply more risk
Need sufficient volume of points (25)
Costs increase as reporting interval decreases
What is current reporting interval?
Hanford PI Forum

9. Creating The Control Chart
Hanford PI Forum

10. The Baseline
The Baseline on a control chart consists of the average (center) line, a three-standard deviation Upper Control Limit (UCL) and a three-standard deviation Lower Control Limit (LCL).
The Baseline allows us to predict the future, and evaluate for trends.
A procedure for choosing the proper form of control chart, and calculating the baseline and control limits is available at
Hanford PI Forum

11. Establishing a Baseline
The goal is to establish one or more baseline time intervals with no trends within each interval.
The “MW Rule” helps to show if a baseline is likely to be “good”.
A “good” baseline detects future trends with a minimum of false alarms.
If a trend is detected, we don’t want it to be due to too few data points in the baseline, causing the baseline to have been inaccurate.
Hanford PI Forum

12. “MW” Rule
3 changes of
Direction
Hanford PI Forum

13. Initial Strategy Use the first 25 points for a trial baseline
If less than 25, use all points, assuming MW rule is met
Evaluate for trends within this trial baseline
NOTE: Do not change a baseline unless it is “proven guilty” by a Trend
Hanford PI Forum

14. This is a Trend! One point outside the control limits
Two out of Three points two standard deviations above/below average
Four out of Five points one standard deviation above/below average
Seven points in a row all above/below average
Ten out of Eleven points in a row all above/below average
Seven points in a row all increasing/decreasing.
Hanford PI Forum

15. Remove any Trends from the Trial Baseline
Do show all data, but change the average and control limit calculations by:
Dropping data off of the beginning
Dropping data off of the end
Dropping individual datum point(s) and circling them
Split into two or more baselines
Hanford PI Forum

16. Initial Baseline
In all cases, trends should be investigated to determine the special cause(s)
A circle or a shift in the baseline is a trend
Generally, short term shifts are left circled, long term shifts have a new baseline
Remember – the goal is Prediction of Future Performance
Hanford PI Forum

17. Three data points removed from the beginning of the data set, leaving a stable baseline from the fourth point onwards:
Hanford PI Forum

18. Three data points removed from the end of the data.
Hanford PI Forum

19. Two data points within the data have been removed, and the remaining data used to construct the baseline.
Hanford PI Forum

20. Both baselines satisfy the MW Rule
The data have been split into two baselines. A permanent decrease in the level appears to have occurred.
Both baselines satisfy the MW Rule
Hanford PI Forum

21. Real Example
Hanford PI Forum

22. Standard Deviation
Traditionally, the average range has been used to calculate standard deviation
I have found that the statistical standard deviation works well and is easier to calculate in Excel
Watch out for outliers inflating the standard deviation, however
Hanford PI Forum

23. Outlier Example
Hanford PI Forum

24. Low Rate Trending If zeroes are causing problems:
Leave runs of 7 or more zeroes out (they will always be below average)
Lengthen the data interval (i.e. from monthly to quarterly)
Try “Low Rate Trending” (
Hanford PI Forum

25. Low Rate Trending Description
Order the data by date of occurrence
Determine the number of days between events
Divide the number of days between events into 365
Plot the resulting rate per year in the date sequence on an x-chart control chart
Hanford PI Forum

26. Low Rate Example
Ref: Understanding Variation The Key to Managing Chaos, Wheeler
Hanford PI Forum

27. Establish Expectations
“Stable” performance is not necessarily good
Management needs to determine if the current stable baseline is “acceptable” or “unacceptable”
May use benchmarks, customer opinion, risk analysis, or management philosophy to make this decision
Hanford PI Forum

28. Monitoring Update charts on the required time interval
Check for trends against the trending rules
Circle any trends, inform owning management and look for special cause(s)
Do not shift a baseline unless there is a trend (baseline proven guilty)
The following slides show an example trend’s life cycle
Hanford PI Forum

29. Initial, Stable Baseline
Hanford PI Forum

30. Birth of the Trend
Hanford PI Forum

31. Initial Actions Report the trend Search for Special Causes
Compare Pareto chart of detailed data during the trend to previous stable time interval
Consider Corrective Action Management necessary steps
Hanford PI Forum

32. Maturity
Hanford PI Forum

33. New Baseline Established
Hanford PI Forum

34. Goal Setting
Remember the hazards of numerical targets and the negative behaviors that result (Red Bead Experiment)
We can use the SPC charts to establish directional goals
If we must, we can establish numerical criteria based upon SPC
Hanford PI Forum

35. Directional Goal Setting
If there is currently an improving trend, the goal is to continue the improvement as long as feasible.
If there is currently an adverse trend, then we want to end the trend and stabilize performance
Consider Corrective Action Management, Lessons Learned communications
Hanford PI Forum

36. Directional Goal Setting
If the chart is currently stable, the following question must be answered:
Is Performance Stable at an Acceptable Level?
Benchmarks, customer interviews, surveys, and internal improvement philosophy can help to answer this question
Consider both the baseline average, and the control limits
Hanford PI Forum

37. Maintain Current Performance
If current performance is acceptable, the goal is:
Maintain current performance without any adverse statistically significant trends
Hanford PI Forum

38. Generate a statistically significant improving trend
Improvement is Needed
If current performance is NOT acceptable, the goal is:
Generate a statistically significant improving trend
Hanford PI Forum

39. But I must have a number!
If you must have a numerical target, say for the next fiscal year, then try this approach
Look at the baseline average and standard deviation of the monthly chart
The standard deviation for 12 months is the monthly standard deviation divided by the square root of 12
Set the fiscal year goal as the baseline average, plus or minus 3 times the 12 month standard deviation
Hanford PI Forum

40. Example
You are stable at an average of 10 per month with a standard deviation of 3
The annual standard deviation is 3 divided by the square root of 12 = 0.866
3 x = 2.60
I expect to finish the year with a monthly average between and 7.40, or between 151 and 89 for the year.
Hanford PI Forum

41. Improvement Example
If you must improve, chose the boundary in the improving direction.
If you must raise the number, the target is more than 151 for the year
If you must decrease the number, the target is less than 89 for the year
These targets represent significant change
Hanford PI Forum

42. Conclusion
We will continue to practice with real data and charts in future PI Forums
These strategies fit well with the Control Chart Dashboard process
Proper trending and actions taken will lead to significantly improved performance
SPC can be tied to goal setting
Hanford PI Forum