P Chart
P Chart
P Chart
Introduction
Attribute data are data based on counts, or the number of times we observe a particular event. The events may be:
the number of nonconforming items; the fraction of nonconforming items; the number of defects; or any other distinct occurrence that is operationally defined. defective or conforming; go or no-go; acceptable or not acceptable; or number of defects per unit.
p Chart. The p chart is used to control the fraction of items with the characteristic. Subgroup sizes in a p chart may remain constant or may vary. A p chart might be used to control defective versus conforming, go versus no-go, or acceptable versus not acceptable.
np Chart. The np chart serves the same function as the p chart except that it is used to control the number rather than the fraction of items with the characteristic and is used only with constant subgroup sizes.
Count Charts
c Chart. A c chart is used to control the number of times a particular characteristic appears in a constant area of opportunity.
For example:
defects per air conditioner; accidents per workweek in a factory; and, deaths per week in a city.
u Chart. A u chart serves the same basic function as a c chart, but it is used when the area of opportunity changes from subgroup to subgroup. For example, we may examine:
varying square footage of paper selected from rolls for blemishes; or, carloads of lumber for damage when the contents of the rail cars varies.
The p-chart for Constant Subgroup Sizes The Centerline and Control Limits
Further study reveals that on both day 1 and day 14 the regular delivery truck operator was absent because of illness. Another employee loaded and drove the delivery truck on those days. That individual had never been instructed in the proper care of the product, which requires special handling and treatment. To solve this problem and eliminate this special cause of variation, management created and implemented a training program using the regular driver's experience for three other employees. Any one of these three employees can now properly fill in and perform satisfactorily. Thus the system has been changed to eliminate this special cause of variation.
After the process has been changed so that special causes of variation have been removed, the out-of-control points are removed from the data. The points are removed from the control chart, and the graph merely skips over them. Removing these points also changes the process average and standard error. Therefore the centerline, control limits, and zone boundaries must be recalculated.
The process now appears to be stable and in a state of statistical control. Management may now look for ways to reduce the overall process average of the number of cracked or broken tiles to raise the usable number of tiles per shipment and effectively increase the process output.
Iterative Reevaluations
It is possible -- and not at all uncommon -that by changing the process, removing points that were out of control, and recomputing the control limits and zone boundaries, points that initially exhibited only common variation will now indicate a lack of control. If and when this happens, the system must again be reevaluated to eliminate the newly revealed special causes of variation.
Subgroup Size
When constructing a p-chart, the subgroup size is much larger than that required for variables control charts. This is because the sample size must be large enough that some nonconforming items are likely to be included in the subgroup. If, for example, a process produces 1.0 percent defectives, sample subgroups of size 10 will only occasionally contain a nonconforming item. As a general rule of thumb, control charts based on classification count data should have sample sizes large enough so that the average count per subgroup is at least 2.00. This allows the A, B, and C zones to be wide enough to provide a reasonable working region into which data points may fall for analysis. This is true for both the p-chart and the np-chart, which we discuss later in this chapter.
Subgroup Frequency
Every process goes through physical cycles, such as shifts and ordering sequences. p-chart and np-chart calculations must be based on a sufficient number of subgroups to encompass all of the cycles of a process to include all possible sources of variation. Subgroup data should be collected at a frequency greater than the frequency at which the process can change. This frequency is determined by a process expert.
Number of Subgroups
As a rule of thumb, the number of subgroups should be at least 25 for pcharts and np-charts.
Boundary between lower zones A and B = Boundary between lower zones B and C = Boundary between upper zones A and B = Boundary between upper zones B and C =
The np-Chart
Classification data can sometimes be more easily understood if the data appear as counts rather than fractions. This is especially true when using attribute control charts to introduce control charting and encountering reluctance by some members of the affected community to deal with fractions rather than whole numbers, such as the number of defects. The quantity np is the number of units in the subgroup with some particular characteristic, such as the number of nonconforming units. Traditionally, npcharts are used only when subgroup sizes are constant. As the information used is the same as for p-charts with constant subgroup sizes, these two charts are interchangeable.
The upper boundary between zones B and C for this example is given by
and the lower boundary between zones B and C is given by
Boundary between upper zones A and B = Boundary between lower zones A and B = The results for this example are:
Count Charts
A defective item is a nonconforming unit. It must be discarded, reworked, returned, sold, scrapped, or downgraded. It is unusable for its intended purpose in its present form. A defect, on the other hand, is an imperfection of some type that is undesirable, although it does not necessarily render the entire good or service unusable. One or more defects may not make an entire good or service defective. For example, we would not scrap or discard a computer, a washing machine, or an air conditioner because of a small scratch in the paint.
c-Charts
Areas of opportunity that are constant in size are easier to manage than those that vary, in much the same way as constant subgroup sizes in a p-chart are easier to manage than those that vary. Constant areas of opportunity might be such things as:
one unit of a particular model of a TV set; a particular type of hospital room; one printed circuit board; one purchase order; one aircraft canopy; five square feet of paper board; five linear feet of wire.
When all conditions for an area of opportunity chart are met, and when the subgroup sizes remain constant, a c-chart is used.
The number of events in an area of opportunity is denoted by c, the count for each area of opportunity. The sequence of successive c values, taken over time, is used to construct the control chart. The centerline for the chart is the average number of events observed. It is calculated as:
A count of 59 is within the control limits, while a count of 60 is beyond the UCL.
Because the actual counts are whole numbers, the observation would fall into zones as follows:
The zones each contain a reasonable number of whole numbers and are close enough in size to be workable.
Zone
Upper A Upper B Upper C Lower C Lower B Lower C 6, 7 4, 5 3 1, 2 0 0
Counts
The process is not in control. Special causes of variation are present. Let us assume that the local operators responsible for the final inspection act so that the special causes of variation for points 9, 10, 14, 19, 20, 22, and 23 are identified and the appropriate corrections are made. The data for points affected by known special causes that have been eliminated are deleted from the data set, and the centerline and control limits are recomputed:
The new limits are so close to the old limits that the old limits are used for the next 24 machines produced.
The first five data points for these next 24 machines are well below the lower control limit. Investigation by the local operators reveals that a substitute for the regular inspector counted the defects on those five machines. The substitute was not properly trained and did not identify all the defects correctly. The operators informed management, and management made appropriate changes in policy so that this situation would not recur. These points can now be eliminated from the data. Beginning with machine number 30, all counts are below the process average. Local operators decided that the process has been changed, so a revised control chart is constructed beginning with point number 30.
Note that as the number of labor-hours per month remains constant, the area of opportunity is considered constant month-to-month.
C Chart of Accidents
4 UCL=4
Sample Count
_ C=1
LCL=0
As there are no indications of any special variation, we can conclude that the process is stable and in a state of statistical control. The company may not realize it, but it is in the business of producing accidents at the stable rate of one per month. It will continue to do so until some effort is made to change the underlying process. If no change in the process is made, accidents will continue to be produced at this rate. Consequently, the "SAFETY IS BETTER THAN COMPENSATION" sign is unfair: employees are not empowered to make system changes that would lower the average number of accidents per month; the sign subtly and unjustly shifts the burden of responsibility for safety from management to the employees.
Name of Sales Person Allan Fred Mark David John Phil Total
Examining the above data from the perspective of the System of Profound Knowledge and statistical thinking results in the following c-chart statistics:
According to the above calculations, the number of new accounts generated by a salesman in a six month period can be predicted to be between 13 and 45 new accounts. This large amount of variation is attributable to the sales system. All salesmen are in the same sales system, and they all deserve the percentage pay raise. No one should be rewarded, and no one should be punished. Sales management should focus their attention on improvement of he sales system, not on rewarding and punishing salesmen.
If Fred had generated 66 new accounts (instead of 36) the sales systems statistics would be
In this scenario, Fred is outside the sales system on the high side (he is above 51). Investigation by Freds manager led to the realization that Fred had developed a better telephone procedure for screening potential customers. Fred should receive special recognition because he is outside the sales system on the high side and because his efforts provide guidance for improvement for all salesmen all in the sales system.
u-Charts
In some applications the areas of opportunity vary in size. Generally, the construction and interpretation of control charts are easier when the area of opportunity remains constant, but from time to time variations may be unavoidable. For example, samples taken from a roll of paper may need to be manually torn from rolls, so that the areas sampled -the areas of opportunity -- will vary.
The characteristic used for the control chart, u, is the ratio of the number of events to the area of opportunity in which the events may occur. For observation i, we call the number of events (such as imperfections) the observed ci, and the area of opportunity ai. Thus, ui is the ratio for each point.
The control limits are different for each of the subgroups and must be computed individually for each subgroup
No points indicate a lack of control, so there is no reason to believe that any special variation is present. If sources of special variation were detected, we would proceed as we did with the c-chart - that is, we would identify the source or sources of the special variation, eliminate them from the system if detrimental, or incorporate them into the system if beneficial; drop the data points from the data set; and reconstruct and reanalyze the control chart.
This implies inspecting all of the items and sorting those that conform to some specification from those that don't. Not only is this inspection costly, but it's equivalent to accepting the fact that the process is producing a constant fraction of its output as defective and will continue to do so. Hence, attribute control charts are limited in terms of the level of process improvement they make possible. Additional process improvement is possible with variables control charts.