You can consider to break down the subgroups (for example, split the day in 4 and take 4 subgroups of 50 instead of 1 of 200). To create a pie chart using raw data: Open the data file in Minitab From the tool bar, select Graph > Pie Chart. If the subgrouping period is too long (due to a big sample size) your chart may have very few points even after a considerable elapsed time. To make a chart you need several subgroups (typically 25 or more). If you find a n out-of-control signal after a whole day there would be a lot of scrap that could have been prevented and it will be also difficult to investigate the special cause and to take the appropiate corrective measures. Right-click any of the points, go to Add -> Data Labels 3) Then select Use labels from column, in the text box enter the column with the labels and then press OK. – Subgrouping period: Ideally one want to have a fast feedback. 2) Add the labels to the points using the labels column. However problems may arise from big samples: From a purelly statistical point of view, the bigger the better. In general, you have the smallest size for technical matters of the control chart. The normal is.Ībout the sample size, you will find a discussion also in page 93. It is not stated that it uses the normal approximation to the binomial, but in page 95 you will find the formulas for the control limits, which happen to be “average ± 3 sigmas”, what would make no sense unless the distribution was (or could be approached as) symmetrical. On page 93 you will find the “more than 5” rule. It’s a reference itself and links you to a lot of other references too. Click the paintbrush icon on the right side of the chart and change the color scheme of the pie chart. Click the + button on the right side of the chart and click the check box next to Data Labels. Statistical Process Control (Ford, Chrysler, General Motors), normally referred as AIAG’s SPC Handbook (is pretty inexpensive and can be bought trough the AIAG’s web site). Click the legend at the bottom and press Delete.