GETTING FAMILIAR WITH SPSS
SPSS allows you to analyzed large databases with relative ease. We’ll start
getting familiar with the program by analyzing a simple database that comes with
the program. When we move to the lab, follow the steps outlined below to arrive
at the correct answers.
1. Open SPSS 10.0
2. Select the Open Existing Data Sources option.
3. Click on World 95. This is a subset of a United Nations data set.
4. Data file includes information on: country, population, density, %
urban, predominant religion, life expectance for females and males,
literacy, rate of population increase, infant mortality, average daily
calories, aids cases, birth rate, death rate, aids rate, fertility, literacy
for males and female, climate, and others.
5. Let’s start by applying some descriptive statistics (mean, median,
mode, std. Deviation) to some variables. To get descriptive statistics
follow these steps:
a. Go to Analyze – on the top bar.
b. Go to Descriptive Statistics.
c. Click on Frequencies.
d. Highlight and click on the variables you want to analyze.
e. On the bottom of the box – click on Statistics and select mean,
median, mode, and std. Deviation.
f. Click on Continue
g. Click on OK
6. Answer the following questions:
a. What is the average life expectancy for females: ____ mean
____ median
____ mode
____ std. Deviation
b. What is the primary religion in the largest number of
countries ?
Religion _____________________ Percentage _____
c. What is the second highest?
Religion _____________________ Percentage _____
d. What is the third highest?
e. Religion _____________________ Percentage _____
7. There are several inferential statistics you can you to compare
differences between different groups of people. Today, we’ll examine the
use of Analysis of Variance. Follow the following steps:
a. Go to Analyze.
b. Click on Compare Means.
c. Select an independent variable: religion
d. Select the following dependent variables: infant mortality rate,
population increase, aids rate, % of females who can read.
e. Click on options – then on ANOVA – then Continue
f. Are there significant differences between religious groups in
terms of:
1. infant mortality rate________________________
2. population increase ________________________
3. aids rate ______________________________
4. % of females who can read _________________________
BUS 3542/AUTUMN 2000
LEARNING TO USE SPSS – 2nd EXERCISE
THE OUTLOOK DATA SET
1. Open up the SPSS program.
2. Open existing data set - look for dataset in SBE courses in Faith –
BUS 3542 – datasets – Outlook data
3. Click on Analyze – run a frequency distribution of the data set.
a. What problems do you see with the data set? __________________________
________________________________________________________________
b. What would you do to fix these problems?___________________________
________________________________________________________________
4. Getting to use the Multiple Response Function:
a. Click on Analyze – go to Multiple Response Function
b. Define Sets – Highlight all Problems with Outlook questions, Q8
c. Define Categories – 1 to 25
d. Name new variable – Problem
e. Label – Problems with Outlook
f. Add, then Close
g. Now when you click on Multiple Response you get 2 more options –
frequencies and crosstabs
h. Run frequencies on your new multiple response variable – what
were the top three complaints regarding outlook? What % of respondents
had these complaints?
___________________________
___________________________
___________________________
5. Getting to use Crosstabs:
a. Click on Analyze – go to Descriptive Statistics – the click on
Crosstabs
b. Select Overall Satisfaction with Outlook as the Column variable.
c. Select Type of Student as the Row variable.
d. At the bottom of the box, click on Cells, then indicate you want
Column percents.
e. Again at the bottom of the box, click on Statistics, and check Chi
Square.
f. Click OK.
g. How do the four student groups differ in their satisfaction with
Outlook?
____________________________________________________________
____________________________________________________________
h. Are the differences between groups of students statistically
significant? If you were deciding what training programs to implement,
is there a specific group of students you would focus on? Why or why
not?
______________________________________________________________
6. Examine the means of the questions measuring student satisfaction with
different features of Outlook on a scale from 0-10. Do any of these indicate
cause for action (providing training)?
______________________________________________________________________________________
______________________________________________________________________________________
7. If you were going to target your training to specific groups of
students (f,s,j,s), are there specific features of Outlook that you would
focus on because a given group of students appeared less knowledgeable in
that area? What features, which groups?
___________________________________________________________________________________
___________________________________________________________________________________
___________________________________________________________________________________
___________________________________________________________________________________
USING SPSS
DETERMINING THE MARKETING IMPLICATIONS OF
THE VISA CARD SURVEY
The questionnaire for this in-class assignment is found on the back of your
textbook, Appendix 1, page A-25. The data which will be analyzed using SPSS is
on the Faith server under BUS 3542 –datasets – visa survey. This is actually
a real data set, so it presents some of the complexities of analyzing real data.
The most important complexity is with regard to recoding data. You will notice
that a lot of the questions have a DK category that holds a point value and is
included in calculating means. To eliminate these DKs from the calculations of
means, you will need to recode the data by following a specific procedure.
Recoding Data
1. Click on Transform
2. Click on Recode
3. Select into Same Variables – that’s a much easier procedure than Into
Different Variables, but you run the risk of screwing up a data set if you are
recoding the only copy of your data.
4. Highlight and doubleclick on all similar statements. For example, all the
Likert items on q. 1 need to have the same recodes.
5. Click on Old and New Values
6. Since you want to get rid of the 6 on q. 1, type in 6 under Old Value.
7. Click on System Missing for New Value
8. Click on Add, then click on Continue, then OK on the next screen – and
your data on these questions have been recoded. You need to go through the same
procedure for all the scales where you want to calculate a mean. The number that
you will assign to Missing Values will depend on the specific question, for
example for q. 25 you need to recode number 101 into a missing value, and for q.
32, you need to recode 11 into a missing value.
SELECT A PARTNER. YOU AND YOUR PARTNER SHOULD BE PREPARED TO ANSWER THE
FOLLOWING QUESTIONS IN CLASS ON MONDAY.
a. How likely are teachers to apply for a VISA card with the features
indicated ? (question 25) Note: Unfortunately, information on q. 24 is missing
from the data set, so we don’t know the differences with regard to the 4
conditions specified on q. 24.
b. Which of the features mentioned in question 29 are more important to these
teachers? Why?
c. How would you use the data on question 30?
d. How would you use the data on question 32? What credit card features do
you think are especially attractive to teachers?
e. The questionnaire starts with a Likert Scale. I don’t know the exact
reason for including the statements in this questionnaire, but I assume that
they are meant to measure psychographic dimensions which can be integrated in an
advertising campaign. Which of these statements are important in marketing this
new VISA card to teachers? Are there any themes you would integrate in the
advertising campaign? Why?
f. Is there a specific type of teacher who would be a better target for this
VISA card? Why?
