Minitab: An Introduction

What is Minitab?

Minitab is a statistical program with a spreadsheet-like data worksheet. It is capable of manipulating and transforming this data and can produce graphical and numerical summaries. Minitab also allows you to perform a wide variety of statistical computations. Accessing Minitab's commands can be done through menus and dialog boxes. If you prefer, an interactive command language is supplied. These two forms of command syntax can be interwoven as desired. Minitab programs, however, consist of macros written by the user.

Limitations

This is the student version of Minitab and it has certain limitations. The most prominent of these is the lack of any matrix operations. Secondly, you are restricted in the size of data with which you can work. It should be noted that the student edition is limited to 3500 data points and 100 data columns. Also, ANCOVA and multivariate methods are not supported.

Getting Started

Both the PCs and the Macs support Minitab. With few exceptions, they run very similarly on the two machines. Starting the program, however, is considerably different.
IBM PC:
Login as normal and at the N:\MSDOS> prompt, type "minitab" and press ENTER.
Macintosh:
Login as normal and double-click on the "Applications" icon. In this new folder, double-click on the "Stats Packages" icon. Once this folder opens, double-click on the "Minitab Student" folder and then on the "Minitab Student" icon.

Minitab Menus

Following is a list and brief description of Minitab's menu items. For most tasks these menus will be sufficient as it is generally considered easier to use the menus than to use the programming language.
File
Allows you general file manipulations as well as importing/exporting ASCII data files.
Edit
This menu is responsible for data entry and revisions. Also provided is general cut and paste commands.
Calc
Allows you to generate random data and view data with various probability distributions. Various numerical functions are provided to help with computations. Allows sorting and ranking columns of data. This menu also allows you to perform various row and column operations such as; deleting rows and columns of data, erasing of variables, copying data, stacking and unstacking operations and several others.
Stat
This is the basic statistical calculation menu. It allows you to do basic statistics, multiple and stepwise regressions, various types of variance analysis and nonparametrics. Another feature of the stat menu allows you to display charts.
Graph
This menu, as it maintains, lets the user present various types of graphs. Some types of graphs include; histogram, boxplot, dotplot, stem-and-leaf, scatter plot, time series plot, 3-D plot.

Worksheet Data Screen

Minitab provides you with a spreadsheet-like data screen. It also provides you with some elementary spreadsheet data manipulation routines. Moving around in data sheets is as simple as using your mouse. Entering data is similar to that of most spreadsheets. Data manipulation within the data screen can be accomplished in two primary ways, first by direct manipulation by the user, and second by using the F10 key or clicking on the F10=Menu box displayed in the lower right hand corner of the screen. To save the data screen use the F10 menu to return to your Minitab session or press the ALT-M keys. Then go under the file menu and select "Save Worksheet As". Type your filename and the data is saved.

Some Examples

Entering Data
As previously suggested, data entry in Minitab is very similar to that of a spreadsheet. To enter data into a cell, use the mouse to click on and highlight that cell and then enter the data. Similar data (i.e. - heights of 20 people) are entered in columns with the top cell in the column set aside for a title or description of that column.
Loading Data
For large amounts of data, there is usually a file which supplies the data to save the time of having to type it all in. To load data in this way, simply click on the menu item "File" and click on "Open Worksheet". You might have to change the folder to get at the data you need. (i.e. - it might be in a file inside the "samples" folder.)
Minitab has the option of creating various types of data. This can be done in several ways. If you are in session (not worksheet) mode, you can type in the commands for a set of random numbers. Say you wanted data on 200 rolls of a six-sided die. This would be entered as follows: 
MTB > RAND 200 C1;	(This denotes 200 random numbers placed in
 			column C1)
SUBC > INTE 1 6.	(This denotes integers from 1 to 6)

It is important to include all punctuation (semi-colons and periods) or the computer will not accept the command.
Another way to do this would be by using the menus. Under the menu item "Calc", you can choose "Random Data". Move the mouse right to get another menu and select "Integer". You will then see a form you can move around with either the mouse or the TAB key. In the above case, we would put 200 under the number of rows, C1 under the columns and 1 and 6 as the minimum and maximum values.

Manipulating Data

Now that you have some data, what can you do with it? There are several options if you browse through the menus. Say we want to know some basic statistics about our 200 numbers like the mean, standard deviation, etc. Here is what we do. First, click on the menu item "Stat", then go down to "Basic Statistics". Go right with the mouse and click on "Descriptive Statistics...". Then type the appropriate column in the "Variables" box (C1 in this case) and press ENTER. The session window will come up giving information on row C1.

Following now is an example of the above.

Sample Session:

Say that we have the following data: 
 C1-A   C2    C3   C4     C5
 Name Test1 Test2 Test3 Average
 Bob    75    69   80   74.6667
 Carl   82    86   81   83.0000
 Dean   36    61   55   50.6667
 Fred   29    47   56   44.0000
 Gary   90    95   88   91.0000
 John   91   100   93   94.6667
 Louis  86    78   81   81.6667
 Paul   70    72   61   67.6667
 Roger  69    59   73   67.0000
 Steve  58    68   70   65.3333
If this data were not in a file, we would have to input it manually on the worksheet window. Once the data has been entered, we want to find out some basic information about it. By following the instructions above for "Descriptive Statistics" but entering "c2 c3 c4 c5" instead of "c1", we will flash to the "Session Window" and be shown the following: 
 MTB > Describe c2 c3 c4 c5.

               N      MEAN    MEDIAN      TRMEAN    	STDEV   SEMEAN
Test 1         10    68.60    72.50    	  70.75  	21.68     6.85
Test 2         10    73.50    70.50    	  73.50       	16.54     5.23
Test 3         10    73.80    76.50    	  73.75       	13.17     4.17
Average        10    71.97    71.17    	  72.62       	16.41     5.19

              	 MIN      MAX       Q1       Q3
Test 1      	29.00    91.00    52.50    87.00
Test 2      	47.00   100.00    60.50    88.25
Test 3      	55.00    93.00    59.75    82.75
Average     	44.00    94.67    61.67    85.00
From this, we get some basic statistics. Now say we want a boxplot for the average. This is a simple matter of selecting "Graph" from the menu bar and going down to "Boxplot". As our average has been entered in column C5, we type in "C5" where Minitab asks for the variable and press ENTER.
The result should be something like the following:

Now that we have some of the basics down, we should try something a little more difficult, like an ANOVA.
For this example, we will use some pre-recorded data from a file.
To do this, select "File" from the menu bar and then "Open Worksheet". In the Minitab folder is a file called "Examples". Open this folder and select "DATA". Open this folder and select "GRADES". The worksheet should come up showing 24 people with marks on 3 different tests.
For this example, we are going to test our null hypothesis: 
Ho : m1 = m2 = m3    vs. the Alternate hypothesis
H1 : At least two means differ.
We will test this hypothesis at the .05 level of significance.
To perform this ANOVA, click on the "Stat" menu and select "ANOVA". Move the mouse right and choose "One-Way (Unstacked)". It will ask for the columns involved so type "C3 C4 C5" and press ENTER. The following should appear on your session window. 
MTB > AOVOneway c3 c4 c5.

ANALYSIS OF VARIANCE
SOURCE     DF        SS        MS        F        p
FACTOR      2       641       320     0.95    0.391
ERROR      69     23209       336
TOTAL      71     23850
                                   INDIVIDUAL 95 PCT CI'S FOR MEAN
                                   BASED ON POOLED STDEV
 LEVEL      N      MEAN     STDEV  ---------+---------+---------+-------
TEST1      24     80.00     15.62            (-----------*------------) 
TEST2      24     81.29     13.28              (-----------*------------) 
TEST3      24     74.42     24.26   (-----------*-----------) 
                                   ---------+---------+---------+-------
POOLED STDEV =    18.34                  72.0      78.0      84.0
From this information, we get a relatively low F-value (only 0.95) and our p-value is certainly greater than 0.05 (0.391). Thus we decide to not reject the null hypothesis.

As you may have noticed, commands that are activated through the menus show up on the Session screen. Once you get proficient with Minitab, you may find it faster to type in the commands yourself rather than refer to the menus all the time.

Getting Help
Like many of the Math and Stats programs, Minitab also has a help function. It can either be accessed through the menus or you can ask for help on a specific subject by typing "Help subject" on the Session screen where "subject" could be ANOVA, Regression, etc.

Any problems with or suggestions for this page can be sent to
webmaster@math.usask.ca