Locked lesson.
About this lesson
Minitab is the statistical analysis software application that is most often used with Lean Six Sigma projects. Minitab has a Wizard that guides you through the setup and design of a Design of Experiments study. This lesson demonstrates how to use that Wizard.
Exercise files
Download this lesson’s related exercise files.
DOE in Minitag.docx60 KB DOE in Minitag - Solution.docx
274.5 KB
Quick reference
DOE in Minitab
Statistical software applications are often used when doing Design of Experiments studies. The Minitab application is a popular application often used with Lean Six Sigma projects and it has wizards that will plan and analyze DOE studies.
When to use
The use of a statistical software application to plan and analyze a DOE will simplify the process for the study team. For that reason, the software is especially helpful with large and complex studies.
Instructions
Minitab is a commercially available statistical analysis software application. It is widely used in the Lean Six Sigma community. Minitab provides a free 30 day trial so you can use this application to complete this lesson. There will be additional Minitab lessons in the next section, so carefully plan when you want to start your trial. Minitab upgrades their software to new versions periodically, so the look of the DOE wizards may be slightly different depending upon which version you are using.
The decisions we have discussed in previous lessons with respect to designing a Full Factorial DOE study are decisions that are entered into the Minitab DOE wizard. Minitab will then create a study worksheet that identifies the control factor configuration for each run and that is where the results are recorded. Once the results are recorded, Minitab will be able to conduct analyses of the data, but that will be discussed in a later lesson.
Minitab works like standard Microsoft products, in particular, Excel. To conduct a DOE, start by selecting the Stat pull down menu at the top of the screen. Select DOE from the menu and then select Factorial and Create Factorial Design from the succeeding menus.
In the wizard that appears, I normally select the default 2-level factorial and the number of control factors.
Next select the Design button on the wizard panel. This brings up another panel where you can select Full Factorial design and select the amount of replicates, center points and blocks that you want. When finished click OK to exit that panel.
Now on the DOE design wizard panel select the button for factors. This brings up a panel where you can name each factor, select the factor type and set the low and high levels for each factor. Click OK to exit that panel and then click OK to exit the DOE design wizard.
Minitab will then open a worksheet that has each run of the study listed on a separate row. Minitab will have randomized the runs. Each run will have the control factors for that run designated. Record the response variable data in the first column to the right of the DOE design plan. Label that column with the response variable title. If there are multiple response variables, label a column for each one.
You are now ready to begin testing and recording data.
Hints & tips
- Minitab operates similarly to Excel. If the operators who are conducting the test do not have access to Minitab, copy the Minitab worksheet to an Excel worksheet. Once you have finished data collection, you can copy the response variable column from Excel and paste it into Minitab.
- The Minitab analysis will depend upon the DOE study design. Do not insert rows or columns in the middle of the Minitab worksheet, it will create errors in the analysis.
- Minitab does its analysis based upon a data set organized into columns not rows. If collection run data in Excel, record it in a column so that it can be copied and pasted to Minitab.
- 00:04 Hello, I'm Ray Sheen, and I wanna wrap up this discussion about full factorial
- 00:09 DOE studies by introducing how to use Minitab to setup and plan the study.
- 00:15 Minitab is one of the commonly available software applications that provides
- 00:20 a suite of statistical analysis for use in engineering and operations.
- 00:24 Minitab is the most commonly used statistical application with
- 00:27 Lean Six Sigma projects.
- 00:29 Your organization may use a different statistical analysis application, but
- 00:33 it will probably still have similar features.
- 00:36 If you want to use Minitab with this course,
- 00:38 Minitab provides a 30 day free trial that you can use to complete the exercises.
- 00:44 As a statistical software application,
- 00:46 Minitab is able to do a variety of statistical analysis on a data set.
- 00:51 And when you open Minitab, you will see three areas on your screen.
- 00:55 The menu is at the top and you use this to select your analysis and format your data.
- 01:00 The table at the bottom is where the data will be entered, and the session window in
- 01:04 the middle provide summary of the analysis that have been done on the data.
- 01:08 Some of the analysis can also be graphed and
- 01:10 a separate graph tab will be created when graphs are chosen.
- 01:15 If you are familiar with Excel, the Microsoft spreadsheet application,
- 01:19 you won't have any trouble navigating and using Minitab.
- 01:22 Just like Excel, data is entered in columns and rows.
- 01:26 The menu across the top are similar in both naming and
- 01:29 function as the Excel menu.
- 01:31 In fact, you can even directly copy data from Excel Minitab, or Minitab to Excel.
- 01:37 Let's narrow our focus to the design of experiments using Minitab.
- 01:42 Minitab will not only analyze your DOE data, it will help you design your stud.
- 01:47 You just need to answer the study design questions that we've already discussed.
- 01:51 Start at the Stat pull down menu from the top of the screen, then select DOE.
- 01:57 Now select Factorial and finally, select Create Factorial Design.
- 02:03 Now based on your study objectives and constraints
- 02:06 you will need to answer questions about the type of study, the factors and
- 02:10 levels, randomization, and the use of replicate center points and blocking.
- 02:14 Let's see how this is done.
- 02:16 Minitab is able to plan the details for our study.
- 02:20 When you select Create Factorial Design,
- 02:22 the window on the left side of the screen opens.
- 02:25 Start by selecting the type of factors.
- 02:27 Normally use the default 2 level factors, then enter the number of control factors.
- 02:33 For this example, we'll select six.
- 02:36 Now, to record many of the decisions, you need to select the Design button.
- 02:40 And then the window on the right side shows up.
- 02:43 The first thing is to select the design type,
- 02:46 in our case we've selected Full Factorial.
- 02:49 Now select the number of replicates,
- 02:51 I've chosen two, then select the number of center points, I've chose three.
- 02:55 Then select the number of blocks, again, I've selected two.
- 02:59 If you don't recall to select these aspects to the DOE study design,
- 03:02 go back to the lesson on factors and review the material.
- 03:06 Now you can hit the OK button in the design window.
- 03:09 So now it's time to set up the factors.
- 03:12 Select the Factor button which is next to the Design button
- 03:15 on the Create Factorial Design window.
- 03:17 This brings the factors windows into view and now you can setup each factor.
- 03:22 Give the factor a meaningful name, not factor A, but
- 03:25 things like cycle time or temperature.
- 03:28 Then select whether the factor is quantitative or qualitative.
- 03:31 If quantitative, select numeric, for qualitative select text.
- 03:36 Finally, put in the actual values or settings that you will use for
- 03:40 your high and low levels.
- 03:42 Now, click OK on this panel and the first DOE design panel.
- 03:47 When you've clicked OK on all the design panels,
- 03:49 Minitab will create your run table.
- 03:53 As you can see, this table has organized and planned your DOE study.
- 03:57 The first column is the order of the runs
- 04:00 if you are using the Yates numbering method, but
- 04:03 the actual order was randomized and that's in column number two.
- 04:07 And you can see in the next two columns which runs are center point and
- 04:11 which are blocks.
- 04:12 The reason we have so
- 04:13 many center points here is because we had several qualitative factors.
- 04:18 And whenever you have one of those it doubles the number of center points,
- 04:22 with two of them it actually doubles that number again.
- 04:25 Then there are the columns that show the control factor settings for each run.
- 04:31 The last thing you must do is to label the first empty column
- 04:34 with the name of your response variable that you'll be measuring on each run.
- 04:38 In my case it is a poll test, if you have multiple response variables,
- 04:43 just create a column for each one.
- 04:45 You are now ready to run the tests and record the data.
- 04:49 We'll talk more about interpreting Minitab results in a later lesson.
- 04:54 Minitab is an excellent statistical analysis tool for doing DOE studies.
- 04:59 It has automated many of the design and
- 05:01 setup steps that can help you to manage your DOE study plan.
Lesson notes are only available for subscribers.
PMI, PMP, CAPM and PMBOK are registered marks of the Project Management Institute, Inc.