You are here:Home>Learning Center>User's Guide>Chapter 2 - Basic Functions>Chapter 2 – Section 5

User's guide chapter 2 section 5 & 6



Previous: Chapter 1 Next: Chapter 3

2.5 – SPC Charts

Statistical Process Control (SPC) charts support your quality and process engineering. You can use them to identify key areas for improvement in a process, and then to monitor process improvement over time. They present complex data in easy to understand formats.

There are several types of SPC charts: Histograms, Run charts (trend charts), Pareto charts, Control charts, Scatter charts, and Pie charts. They are all valuable as planning tools. The next two parts of this section will focus on Pareto charts as an example of how you can use SPC analysis to improve your business planning. The final two parts of this section discuss the use of Control charts to track variation in performance.

2.5.1 – Working with SPC Charts

To select and place SPC charts in a diagram, use the DataAnalyzer Chart wizard on the Insert menu. Select which SPC chart you want to use, and then type your information into the active spreadsheet in the diagram space. DataAnalyzer will graphically convert the data for display in the diagram.

2.5.2 – Using Pareto Charts to Prioritize Process Improvement Efforts

Pareto charts allow you to gather data about the problems you are examining and present graphic representation of the frequency or size of each problem. This provides a way to identify those problems that offer the greatest opportunities for improvement.

The Pareto chart is based on the idea known as the 80-20 Rule: that 80 percent of any result can be attributed to 20 percent of the activities. For example, of 100 errors, 80 can be eliminated by correcting only 20 percent of the causes. The Pareto chart allows you to present your cause data so you can prioritize the causes on which you will focus. It also guides you away from choosing solutions that will worsen the existing situation.

A Pareto chart gives you a different type of information than other ProcessModel analyses provide. Typically, a ProcessModel identifies 1) processing bottlenecks, 2) interdependency problems, 3) resource constraints, 4) cycle time problems, and 5) non-determinate processes. In contrast, a Pareto chart 1) identifies how significant each contributing cause is in a situation, and 2) gives you a visual format to show that relative significance to other people.

Pareto Charts

Analyzing a Situation with a Pareto chart

The basic steps for analyzing a situation with a Pareto chart are:

1. Identify the general problem area you want to investigate and, within that problem, select the specific issues you will analyze.

2. Choose the most useful unit of measurement (such as frequency or cost) for your data collection.

3. Choose the time period for your data collection.

4. Gather current data or review historical data.

5. Compare the relative frequency or cost of each problem category.

6. Set up a Pareto graph by placing the problem categories on the horizontal line and the frequencies on the vertical line. Include the unit of measurement in the chart labels.

7. (Optional) Draw a cumulative percentage line showing the portion of the total that each problem category represents.

8. Interpret the results. Ask: What factor has the most impact on the goals of our business and customers? Remember that the most frequent or most expensive factor is not always the most important.

Variations of the basic Pareto chart can provide additional information by displaying the basic data in different ways. Frequently used variations include:

Major Cause Breakdowns—Break down the tallest bar (i.e. often the cause with greatest effect) into subcauses in a second, linked Pareto.

Before and After Comparisons—Draw “new” Pareto bars side by side with the original Pareto to show the effect of a change. Present the comparison as one single chart or two separate charts.

Data from Multiple Sources—In side-by-side Pareto Charts, show data you have collected on the same problem but from different departments, locations, equipment, and so on.

Data in Different Measurement Scales—Use the same categories, but measure your results differently. For example, compare the measurement of cost and frequency for the same categories.

2.5.3 – Using a Pareto Chart: A Tutorial

Step 1: Setting up the Chart

How To – Select Pareto Chart

1. Click the File menu, and point to New Process.

2. On the Insert menu, click the SPC Chart button. The Chart wizard will appear.

3. On the wizard, click Next to view a list of chart types.

4. Select the Pareto chart from the list in the Chart wizard.

5. Click Finish. A blank data chart will appear.

Pareto Charts

Step 2: Putting Data into the Spreadsheet

How To – Enter data and adjust labels

1. Notice the tabs in the lower left corner. The Spreadsheet tab lets you view and modify the spreadsheet data. The Chart tab lets you view and modify the chart.

2. Double-click the column heading Label. The Column Header Text dialog box opens.

3. Type Process Step, and click OK. Notice the column is not wide enough to display the entire heading.

4. Point to the line between the two column headings. A horizontal double headed arrow appears. Horizontal double headed arrow

5. Drag the line to increase the width of the column until you can see the entire heading.

Adding data to spreadsheet

6. Label the other column Idle Time (hours).

7. Type in the names of the process steps and the idle times in hours.

Adding data to spreadsheet

Step 3: Displaying Data in a Pareto Chart

How To – Display Pareto chart

1. Once you have the Idle Time information and Step Labels in the spreadsheet, click Chart at the bottom of the spreadsheet. A Pareto chart appears, displaying the data you entered.

Display Pareto chart

2. The following results are shown:

The “Receive Approved Purchase Requisition from Manager” step appears first and displays at the far left position. It shows five hours of Idle Time. The next-largest Idle Time step is the “Interoffice Mail Form to Manager” step.

The vertical line at the left of the chart displays the amount of Idle Time. The cumulative frequency (up to 100 percent) is the vertical line at the right of the chart.

The Pareto Chart also displays a curve that identifies what portion of the total causes are attributed to each individual cause.

Important information to be aware of If you wish to change the color of the bars, the style of the background, or the placement of the legend on the chart, follow the procedures in Change the chart appearance.

Important information to be aware of To delete the chart, follow the procedure in Delete a chart.

Step 4: Using Pareto Information to Improve your Processes

Compare the relative contribution of each factor to the results you are observing in the model.

The chart illustrates that the greatest amount of idle time results from two causes:
1) staff must wait to receive the manager’s approval of their purchase requisitions, and
2) staff does not have an interoffice mail form to submit purchase information quickly to the manager. It is likely that if you can solve the problems in these two areas, you can resolve 55 percent of all the reasons for customer calls about slow service. Therefore, the Pareto chart has given you valuable information on how to focus your business resources.

2.5.4 – Pareto Chart: Basic Procedures

How To – Insert a Pareto chart

1. Click the File menu, and point to New.

2. In the Zoom box, click 100%.

3. On the Insert menu, click SPC Chart. The Chart wizard opens.

4. Click Next. A list of SPC charts displays. Pareto Chart is already highlighted.

5. Click Finish. A spreadsheet opens where you enter the data to be charted.

How To – Label columns

Use the chart you created in Insert a Pareto chart.

1. Notice the tabs in the lower left corner. The Spreadsheet tab lets you view and modify the spreadsheet data. The Chart tab lets you view and modify the chart.

2. Double-click the column heading Label. The Column Header Text dialog box opens.

3. Type a label for the first column (e.g. Process Step), and click OK.

4. If the column is not wide enough to display the entire heading, point to the line between the two column headings. A horizontal double-headed arrow appears. Horizontal double headed arrow

5. Drag the line to increase the width of the column until you can see the entire heading.

Label columns in spreadsheet

6. Type the label Idle Time (hours) for the other column.

How To – Format the chart

1. Click the Chart tab. The chart appears. No data is in the chart unless you are using the chart created in the previous section.

2. Click the right mouse button on the heading, Pareto Chart, and then click Format Title.

3. If needed, drag to highlight the text in the Text box, type Purchase Requisitions, and then click OK.

Important information to be aware of You can reposition the heading by left clicking on it and dragging it to the desired location on the chart.

4. Click the right mouse button on the y axis label, # of errors, and then click Format Title.

5. Drag to highlight the text in the Text box, type Idle Time (hours), and then click OK.

6. Click the right mouse button on the bottom line, the x axis, and then click Insert Titles.

7. Select Category (X) Axis, and then click OK.

8. Click the right mouse button on the x axis label, and then click Format Title.

9. Drag to highlight the text in the Text box, type Process Step, and then click OK.

Format the chart

10. Drag the chart toward the title to leave more room at the bottom for the process step labels.

11. Point to the bottom handle of the hatched area, and drag downward to increase the height of the chart.

Format the chart

How To – Add data

Use the chart you formatted in Format the chart.

1. On the View menu, click the Spreadsheet tab.

2. Click the first cell in the Process Step column.

3. Type the following information in the appropriate columns:

Add data to chart

4. Click the Chart tab. The chart data appears on the chart.

Important information to be aware of You may also use the Data Import Wizard tool from the Tools menu to import an existing data file.

How To – Change the chart appearance

Use the chart you created in Step 2: Putting Data into the Spreadsheet.

1. Click the right mouse button on one of the red bars. The bars on the chart are called the series.

2. Click Format Series.

3. Click the Fill Color box, click blue, and then click OK. The bars change to blue.

4. Click the right mouse button to the side of the chart, and then click Format Chart.

Important information to be aware of Stay inside the hatched lines while working with the chart.

5. Click the Backdrop tab.

6. Click Gradient, click the To Color box, and click yellow.

7. Click OK.

8. Drag the legend to the lower right corner.

9. Click the right mouse button beside the chart, and then click View in 3D.

10. Click outside the hatched lines to place the final chart on the page. To make changes to the chart, right click and select Micrografx SPC Chart Object.

11. On the File menu, click Close. Click No when prompted to save the chart.

How To – Delete a chart

Use the chart you created in Add data.

1. Click the Selector tool on the Toolbar.

2. Click the SPC chart.

3. Press DELETE.

4. On the File menu, click Close. Click No when prompted to save the chart.

2.5.5 – Using a Control Chart to Track Variation in Performance

A control chart allows you to track consistency of performance within an identified range of values. Variations within the range are assumed to be normal, whereas variations outside the range alert you to some special problem that should be investigated. Control charts are complex statistical tools. You should refer to detailed sources for additional information. This section describes how you can use ProcessModel to create a control chart.

Control Chart to Track Variation

Control charts operate on the premise that all processes have some inherent variation. You should pay attention to extreme variances from the average (called special causes), but normal variances from the average (called common causes) should be ignored in the short term and improved through a long-term process improvement effort.

2.5.6 – Using a Control Chart: A Tutorial

To use our procurement example again, let’s say you want to track the average cost over time of processing a purchase requisition.

Step 1: Gathering the Data to Measure

Gather your information about the costs of all purchases within a period of time; for example, one week. Average the costs and record this information, as well as the highest and lowest cost for that period. We will use the data shown in Step 2: Creating the Control Chart. We will plot it on an X Avg, R (X Average, Range) Control Chart by using the Chart wizard.

Step 2: Creating the Control Chart

How To – Create the Control Chart

1. On the insert menu, click the SPC Chart button. The Chart wizard will appear.

2. On the wizard, click Next to view a list of chart types.

3. Select the X Avg, R control chart.

4. Enter the data into the spreadsheet.

Control Chart

5. To display the data as a control chart, click Chart. ProcessModel defines the Upper Control Limit (UCL) and Lower Control Limit (LCL) and displays the information on two charts.

The top chart displays the average values within the UCL and LCL. The average values do not vary as much as the range values do, since these values are averaged numbers of all the values for a period of time.

Control Chart to Track Variation

The bottom chart displays the range of values. Notice that the values in this chart vary widely. These values vary so much more than the top chart because the bottom chart shows the difference between the highest and lowest of all the values.

Control Chart to Track Variation

Both charts are helpful when trying to determine if the process is in statistical control.

Step 3: Evaluating the Information

The Control chart indicates the range of variability over time. A high degree of variability may indicate that your process can be improved. Range values outside the UCL and LCL also indicate a special-causes situation. Research and address any special causes that are outside the boundary of the UCL or LCL.

These charts indicate something happened during week 25 that resulted in a loss of statistical control in the process.

Evaluating the Information on chart

Important information to be aware of Evaluation Applied to a Process When you apply the control chart to a process, look for occurrences outside the UCL or LCL. These occurrences indicate that the process is “out of control.” Over time, efforts to improve the process should reduce the distance between the UCL and LCL, indicating an improvement in process stability.

Important information to be aware of Evaluation Applied to a Maintenance Program You may w ant to track gas mileage performance of your car or a fleet of vehicles. Set up a form to gather gas mileage information that you need. With ProcessModel’s control chart, you can check that the data falls within range. When the data falls out of range, you may check whether the truck needs a tune-up, air in its tires, or other maintenance. This process also helps identify a regular maintenance program that may be justifiable, since it can reduce gasoline costs

This variation in week 25 is probably an area where you will want to focus further analysis. Identifying the cause of the variation could lead to a revision of your process and a return to process stability.

Step 4: Ending the Control Chart Tutorial

How To – End the Control Chart Tutorial

1. On the File menu, click Close.

Click No when prompted to save the chart.

Previous: Chapter 1 Next: Chapter 3