At this point you have generated product data using a program based on the real production line. The quality control engineer wants to know the characteristics of the product stream, the shape of the data, its location and variability, how well specifications are being met, and the possibilities for reducing variability, scrap, and rework. An operator would be interested in monitoring production to avoid over and under adjustment.

Now it is up to you to analyze the data to reduce variation, respond to problems, and meet specifications. You may create reports characterizing current production and comparing it to standards or to past production.

An instructor for a statistics course will assign students to analyze production data using the statistical methods being taught. The pages in working with the data give examples of how the production data might be used in a course.

All of these questions can be addressed using common statistical methods. The examples in this guide use Microsoft Excel because it is commonly available. Other statistical and charting software would work as well or better.
Applications, page 1
Using the simulation default settings, a run of 8640 10.0 inch blades yielded 30 samples. The sample size was set at 5. Figure 13 shows the hardness data pasted from the computer clipboard into an Excel spreadsheet. Columns C through G contain the individual hardness measurements. Each row contains one sample with the date and time it was taken. Sample measurements were taken from each batch after tempering.

1. Determine the shape of the distribution of individuals by making a histogram of the measurements. Other issues: How were the samples chosen? Do samples of 5 from each batch constitute a random sample? Do they constitute a simple random sample?

2. What are the basic descriptive statistics for this run?

3. Make x bar and s charts based on the 30 samples, using Excel, other software, or hand charting. Is the process in statistical control?

4. Produce another 30 batches of the same size blade. Monitor the production for out of control conditions by continuing the chart above.

5. Determine the cpk ratio of the current production using the specifications contained in the work instructions.

6. What are the causes of variation in hardness? Which causes can be detected and reduced ?

7. Answer the same questions for the length data.
Figure 13. Hardness data pasted into a spreadsheet.

The following pages in Working with the data show how the example questions above can be answered.
Back Statistics and SPC: 1 2 3 4 5 6 7
Process Simulation Home Page