An apology/disclaimer for Mac users: Everything I describe on these Minitab assignments will refer to the PC version of Minitab. If you want to try to use the Macintosh version, some of the things I say may not jibe; thus, use Mac at your own risk for these assignments.
 |
A Minitab worksheet consists of a table of values in which each column
is a variable and each row is a case or individual. I have prepared
a worksheet for you to download off of the I drive.
Minitab version 13 is freely available on any CAC PC on campus. Go to your favorite computer lab, log in, and start up Minitab. After you start up Minitab, go to the file menu (shown on the left) and select "Open Worksheet". Note that you cannot use "Open Project" if you just want to open a worksheet; thus, control-o won't work. The file you want is I:\STAT\460\heights.mtw |
If for some reason you are unable to access the I drive or find the file (a distinct possibility this week due to an unexpected problem saving files to the I drive), here is a text version of the dataset. Save it to your computer's hard drive, then use Minitab to open it as a worksheet file of type text.
|
After you open the worksheet correctly, you should see something
like this:
Note that the Minitab screen is split into two parts, the data window at the bottom (partially shown here) and the session window at the top. Once again, note that there are two columns (the variables, namely sex and height) and many rows (the cases; each case is a separate individual). |
 |
I'd like you to get in the habit of enabling the commands in the session window. Enabling the commands gives you a chance to take a look at some of the Minitab commands that are created when you perform procedures using the menus and that actually instruct Minitab to do what it does.
 |
To enable the commands, first you need to make the session window
active (the active window is the one with the blue menu bar). To
do this, simply click anywhere in the session window. Then,
you should be able to select Enable Commands from the Editor menu.
If you've done this correctly, there will be a check mark next to
Enable Commands the next time you look at the Editor menu and the letters
|
 |
Because recoding is a form of data manipulation, go to the Manip menu and select Code, then Numeric to Text. You'll get a dialog box that looks like the one on the left (without any of the blanks filled in). The blanks as they are filled in here instruct Minitab to code the data from column C1 into column C1. Thus, this recoding will actually replace C1 with its revised version. (Note: We later undo this recoding in Part 5. You may therefore want to modify this step slightly if you're feeling confident, but do this at your own risk!) I hope that the other entries in the dialog box are self-explanatory. Type them in, click OK, and see if you understand what happens. |
Now that you've changed C1 from a numeric variable to a text variable, go to the question list and answer questions 1 and 2.
|
After you select Stem and leaf, you should see a dialog box like
the one on the right.
Click on the "C2 Height" line in the leftmost box, then click Select. The name of the Height variable should appear in the Variables list. Notice that you were not given the "C1 Sex" variable as a choice. Think about why (this will be question 4). Click on OK. You should get a stem and leaf plot produced in the session window. |
 |
Now it's time to answer questions 3, 4, and 5.
 |
From the Stat menu, select the Basic Statistics submenu and
then Display Descriptive Statistics. You should get a dialog box
like the one shown at the left.
Put the height variable in the Variables list. Then check the By Variable box and put the sex variable in that blank. If you've done this correctly, after you click OK you should get a bunch of output in the session window telling you all sorts of stuff about Females, Males, and * (missing). As a check, you should see that the mean height is 64.788 for females and 70.623 for males. |
Answer questions 6 and 7 based on your output.
Use the stem and leaf command to produce two stem and leaf plots, one for males and females. To do this, you'll need to recode the sex variable into a numeric categorical variable (this is a shortcoming of Minitab--it should be possible to use a text variable as the By variable) and use that variable as the By variable in the stem and leaf dialog box. I'll leave it to you to figure out how to recode from a text variable to a numeric variable; it's very similar to the recoding in Part 2. Question 8 asks you to copy and paste the female stem and leaf plot from the session window.
|
The stem and leaf plots don't reveal any serious non-normality, so
we may proceed with the t-test.
From the Stat menu, select Basic
Statistics and then 2-sample t. You'll get the dialog box shown
at the right.
In this case, you want to keep "Samples in one column" selected. The Samples variable is C2, the height. The Subscripts variable is the sex variable--though here again, you'll need the numeric categorical variable from the previous step. Again, this is a shortcoming of Minitab: You CAN use a text variable as the subscript variable, but it cannot contain any missing values or you'll get an error. With the numeric variable, Minitab correctly excludes the missing values from the t-test. |
 |
Please use the equal variances assumption for the test; to do this, you'll need to click the appropriate box before you click OK (this forces Minitab to use the procedures of Section 9.3 instead of 9.4). Use the output from your t-test to answer question 9.
Troubleshooting: If you had any trouble with the t-test in Part 5, it may be because of the missing values for the sex variable. One thing to try is simply deleting these 3 cases entirely. Scroll to them in the data window, click on the row number in the left margin, then choose Delete Cells from the Edit menu. Or, you might come up with a more clever way to delete the offending cases, but I'll leave that to you.