Computer Exercise:

Introduction to MINITAB
36-220
Fall 2014

Objectives

1. To learn the rudiments of the MINITAB software environment.

2. To practice interpreting some graphical and numerical summaries of data.

What is MINITAB?

‘Minitab’ (meaning ‘minicomputer tabulator’) is a software package designed with

simple interface especially for teaching purposes, but which is nevertheless optimized for
real statistical analysis.

Why Learn MINITAB?

▪ Reinforces lecture material in a more engaging way.

▪ Encourages group interaction on the material.

▪ Allows you to ‘experiment’ with statistics.

▪ Empowers you with ability to analyze large, real-world datasets.

▪ Prepares you for software packages that you may encounter in your own specialties
(some of these may include: Data Desk, R, SPSS, SAS, Stata, S Plus, TSP).

Scenario Background:

In a study designed to test the safety of various types of vehicles, vehicles were crashed into a wall
at 35 mph with a crash-test dummy strapped in the driver’s seat.

We will use MINITAB to analyze some of the resulting data.

36-220 Intro to Minitab Page 2 of 6
Fall 2014

Part I: Opening MINITAB

1. Log on to any University-maintained Windows machine, using your Andrew account.

2. Open Minitab by either of the following methods:

select: Windows Start Icon  All Programs  MINITAB  Minitab 17

or, instead: You could type “minitab” in the file search window  then hit enter

What You See

When MINITAB opens, you will see a bottom spreadsheet-like area called ‘Worksheet’,
where you type data, and an upper window called ‘Session’, where various output will
appear.

Part II: Downloading MINITAB data from Blackboard

To view the data in MINITAB:

1. Go to the Blackboard page for the class;

2. Under homework 2, click the link labeled “CRASH”
4. Select ‘Open’ when prompted.

MINITAB should open automatically on your desktop.

Understanding the Dataset

The data show a variety of variables measured for 166 individual cars that were crash tested. The
variables in the data are as follows:

Variable Description
HEAD measure of head injuries (higher is more injury)
CHEST chest deceleration, which is a measure of chest injury (higher is more injury)
AIRBAG 1 indicates an airbag, 0 indicates only seatbelts were used
DOORS number of doors (equals 6 for vans and sport utility vehicles)
YEAR year of vehicle
WEIGHT weight of vehicle in lbs

This is the dataset referred to in the first three lectures.

Naming Columns

Notice the data label ‘DOORS’ in column 4 has been left out. It is always good practice to name
your columns of data. Plant the cursor in the cell where that data label should go, and type
‘DOORS’.
36-220 Intro to Minitab Page 3 of 6
Fall 2014

Part III: Producing Graphical displays in MINTAB

Let’s graphically explore the CHEST variable with a histogram.

Histogram

Recall that a histogram is a graphical display of a set of numbers, in which bars are used
to represent how many numbers fall within a certain range.

To produce a histogram of the CHEST variable, do the following:

1. Select Graph  Histogram

2. Select ‘Simple’, and click ‘Ok’

3. In the resulting dialog box, do the following:

(i) Make sure the cursor is planted under ‘Graph Variables’

(ii) In the left-hand window, double-click on the column name, so that it

appears under ‘Graph Variables’:

(iii) Then click ‘Ok’.

The histogram will appear in its own window.

The plot is an example of a data ‘distribution,’ because it shows how the chest deceleration in that
dataset is distributed along the axis.

You will want to keep the graph open to answer the following questions:
36-220 Intro to Minitab Page 4 of 6
Fall 2014

Question # 1:
CHEST deceleration is measured along the x-axis, and is grouped into intervals (called ‘bins’ or
‘classes’), represented by the width of each rectangle. What is the ‘bin size’? [also called the
‘class size’.]

Question # 2 [describing the shape]:

Now, describe the shape of the distribution (modality, symmetry or skewness [including the
direction if there is skewness], and outliers).

Part IV: Producing Summary Statistics in MINTAB

To compute various numerical summaries on the dataset, do the following:

1. Select Stat  Basic Statistics  Display Descriptive Statistics

2. In the resulting dialog box, do the following:

(i) Make sure the cursor is planted under ‘Variables’

(ii) In the left-hand window, double-click on the column name to select it,
so that it appears under ‘Variables’.

(iii) Click ‘Ok’.

In the upper ‘Session’ window, you should see a variety of summary values printed (you can move
or resize the various MINITAB windows as necessary).
36-220 Intro to Minitab Page 5 of 6
Fall 2014

Part V: Exploring C  Q relationship with Side-by-side Boxplots

Now let’s compare the chest deceleration of vehicles with different numbers of doors. Create
side-by-side boxplots of chest deceleration by selecting Graph  Boxplot  With Groups 
Ok, then for ‘graph variables’ select CHEST, and for ‘categorical variables for grouping’ select
DOOR, then ok.

Question # 3:
Does there appear to be a relationship between the number of doors and chest deceleration?
Briefly say what the relationship might be, based on the plot.

Question # 4:
Why do you think this relationship exists? [Hint: Consider the side-by-side boxplot display
shown in lecture 3.]

Part VI: Exploring C  C relationship with Contingency Table and Conditional Percents

Now let’s explore air bag presence by year. For the purpose of this analysis we will consider year
to be categorical.

First, we can get a visual indication of air bag presence by year of car. Select graphbar chart
 values from a table  simple  ok, then for ‘graph variables’ select AIRBAG and for
‘categorical variable’ select YEAR, then ok.

PRINT OUT A COPY OF THE GRAPH FOR YOUR HOMEWORK. [With the graph selected, you
can either select file  print graph, or else you can right click on the graph and then copy  paste
into a new document.]

Question # 5:
What does the plot suggest about prevalence of air bags in cars from 1987 to 1991?
36-220 Intro to Minitab Page 6 of 6
Fall 2014

Now let’s numerically investigate the relationship between air bag presence and year. Fist, make a
contingency table by selecting stat  tables  descriptive statistics, then ‘for rows’ select
AIRBAG, and ‘for columns’ select YEAR, then ok.

Question # 6:
Based on the contingency table, how many cars in the study that were from 1987 had an airbag?
How many cars in the study from 1991 had an airbag?

Question # 7:
To compare air bag presence by year (i.e., to use year to predict air bag), do we want row
conditional percentages, or column conditional percentages? INCLUDE THIS ANSWER IN
YOUR HOMEWORK.

Now, produce the appropriate conditional percentages by selecting edit  edit last dialog (this
will pull up the previous dialog box), click the ‘categorical variables’ button, and then check the
appropriate box based on your answer to question 7.

Question # 8:
Based on the conditional percentages, what percent of cars in the study that were from 1987 had
an airbag? What percent of cars in the study that were from 1991 had an airbag? INCLUDE
THESE ANSWERS IN YOUR HOMEWORK.

You are done with the Intro-to-Minitab.

[Remember to quit MINITAB and log off if you are in a cluster.]