Inputting data into SPSS 17.0
Start
Programs
SPSS 17.0 for Windows
PLEASE READ THIS
ENTIRE PAGE FIRST. Next take a look at the third page (a table of your
data). The second page describes the table of data. Once you have read and reviewed
all three pages, follow the instructions on the bottom of this page.
To Enter Data:
Open SPSS. A spreadsheet
called UNTITLED DATA will open. In the first column insert numbers from 1 to 25
as shown on the data sheet. Clicking within any cell will select it for data
entry. To move around the spreadsheet use the ENTER key to go down, TAB to go
across to the right, and SHIFT TAB to move left. Or, you can use the arrows on
the extended keyboard to move around within the spreadsheet.
To Input Variable Information (example):
Go to the variable
view and click on the first variable name.
Replace var00001 with ID where
it says Name.
Continue to move across the
row and fill in the information for that variable.
Type- the default is numeric. Do not change for this
example.
Width- the default is 8 spaces. Do not change for this
example.
Decimal places- the default is 2. Change to 0 for this
example.
Label- In the "variable label" bar, type your
variable name, in this example: Identification
number
Missing- This is to identify variables that you want treated
as missing. For example, if a respondent had put in not applicable on a survey
item you would want to code it as a missing variable. Leave blank for this
example.
Columns- The default is 8. Do not change for this example.
Align- This changes how the variables appear on your
screen. Do not change for this example.
Measure- There are three different types of levels of
measurement you can choose from - nominal, ordinal, and scale. Choose the
appropriate measure. In this example, choose nominal.
For the categorical
variables, Province, Gender, Ethnicity, and Religion, you will need to define the categories.
For example, variable 2: Province:
·
Under values
-double click on the box labeled: None…
·
type: 1 in value bar
·
type: Alberta in
value label bar
·
Click Add
·
type: 2 in value bar
·
type: British
Columbia in value label bar
·
Click Add
·
Continue
·
OK
Enter all the data and name
all the variables in this manner, according to the description provided. Repeat
this sequence for all the variables.
Data Description
Var1 Respondent's identification number (ID)
Var2 Province the respondents lives in (PROVINCE)
1 Alberta
2 British
Columbia
Var3 Respondent's gender (GENDER)
1 male
2 female
Var4 Respondent's ethnicity (ETHNICITY)
1 Caucasian
2 Black
Var5 Respondent's age (AGE)
Var6 Respondent's religious affiliation (RELIGION)
1 Protestant
2 Catholic
3 Jewish
4 None
5 Other
Var7 Respondent' s mother‘s education - years of schooling (MAEDUC)
Name your data set and save
it the data (either to disk or to your student file)
Note:
Use this data to complete Exercise 1 Part B.
DATA SET
|
|
id
|
province
|
gender
|
ethnicity
|
age
|
religion
|
maeduc
|
|
1
|
1
|
1
|
1
|
1
|
32
|
1
|
16
|
|
2
|
2
|
1
|
2
|
1
|
37
|
2
|
13
|
|
3
|
3
|
1
|
2
|
2
|
72
|
2
|
20
|
|
4
|
4
|
1
|
2
|
1
|
86
|
3
|
12
|
|
5
|
5
|
1
|
1
|
1
|
30
|
1
|
5
|
|
6
|
6
|
1
|
1
|
1
|
32
|
2
|
10
|
|
7
|
7
|
1
|
2
|
2
|
29
|
1
|
18
|
|
8
|
8
|
1
|
1
|
2
|
29
|
1
|
4
|
|
9
|
9
|
1
|
2
|
2
|
53
|
1
|
6
|
|
10
|
10
|
1
|
1
|
2
|
68
|
1
|
9
|
|
11
|
11
|
1
|
2
|
1
|
19
|
2
|
2
|
|
12
|
12
|
1
|
2
|
2
|
43
|
2
|
14
|
|
13
|
13
|
2
|
2
|
2
|
38
|
4
|
12
|
|
14
|
14
|
2
|
1
|
1
|
45
|
2
|
17
|
|
15
|
15
|
2
|
1
|
2
|
24
|
4
|
1
|
|
16
|
16
|
2
|
1
|
1
|
53
|
2
|
3
|
|
17
|
17
|
2
|
2
|
1
|
20
|
2
|
7
|
|
18
|
18
|
2
|
1
|
1
|
27
|
2
|
11
|
|
19
|
19
|
2
|
1
|
1
|
54
|
2
|
8
|
|
20
|
20
|
2
|
2
|
2
|
25
|
1
|
15
|
|
21
|
21
|
2
|
1
|
2
|
20
|
2
|
1
|
|
22
|
22
|
2
|
2
|
2
|
38
|
2
|
7
|
|
23
|
23
|
2
|
1
|
1
|
20
|
2
|
5
|
|
24
|
24
|
2
|
2
|
2
|
34
|
2
|
10
|
|
25
|
25
|
2
|
2
|
1
|
67
|
1
|
19
|
SPSS Lab Exercise 2
Running Frequencies and Descriptives in SPSS 17.0
There are 2 ways to retrieve the data file you saved in
exercise 1:
a)
Enter SPSS. An untitled data sheet will appear. Click
on File. Click on Open. Click on arrow under Drives: Click on a: (or relevant
drive) Under file name, click on the file (.sav). OK.
b)
Click on My Computer icon. Click on or relevant drive.
Click on data file (.sav). SPSS will be initiated and the data will appear.
Open previous data file from Exercise 1.
Task 1
To run frequencies for each variable, with the data
editor open in the data view, go to:
·
Analyze
·
Descriptive Statistics
·
Frequencies
·
Click on the selected variable in the left box
and transfer it to the Variable(s) box by clicking the arrow. Note: You can transfer more than one
variable to the Variable(s) box and run frequencies for all variables at the
same time.
·
In the same window click on Statistics
·
Select minimum, maximum, and range (they may
already be selected as default)
·
Continue
·
In the same window click on Charts…
·
Bar chart (You can also try a histogram with or
without the normal curve, and a pie chart. However, SPSS will only allow you to
select on 1 chart at a time!)
·
Continue
·
OK
If you want you may name your output and save it. The
computer will give the output an ‘.spv’
extension if you are using SPSS 17 or an ‘.spo’ extension if you are using another version of SPSS. This
indicates that your frequencies are saved as an output file.
Run frequencies for the following categorical (i.e.,
discrete) variables: Gender, Ethnicity, Religion, and Province.
Answer the following questions:
What percentage of the sample is female? ____________
What percentage of the sample is Black? ____________
What percentage of the sample is Catholic? ____________
What percentage of the sample is from Alberta? ____________
Task 2
To run Descriptives for each variable go to:
·
Analyze
·
Descriptive Statistics
·
Descriptives
·
Click on the selected variable in the left box
and transfer it to the Variable(s) box by clicking the arrow. You can transfer
more than one variable to the Variable(s) box and run descriptives for all
variables at the same time.
·
In the same window click on Options
·
Select mean, standard deviation, minimum, maximum and range (some may already be
selected as default)
·
Continue
·
OK
If you want to you may name your output and save it.
Run Descriptives for the continuous variables: Age, and
Mother’s Education (maeduc).
Fill in the following table.
|
Variable
|
Mean
|
Standard Deviation
|
Lowest Value
|
Highest Value
|
Range
|
|
Age
|
|
|
|
|
|
|
Mother’s Education
(in years)
|
|
|
|
|
|
Task 3
Now we want to review the process of cutting and pasting from your SPSS output into a Microsoft Word
document.
SPSS output should currently be open on the computer screen
in front of you. Let's copy the
Descriptives table you just had SPSS produce and paste it into a "hypothetical" Microsoft Word document.
In order to copy and
paste, you must:
1. Go
back to the bottom left corner of your computer screen to the command Start. Click on Start. Click on Programs.
Find Microsoft Word. Click on Microsoft Word. The Microsoft Word
program should open up onto your computer screen.
2. Sometimes
the Microsoft Word program will now
ask you what you want to do. If the program asks, you want to create a new document.
3. Minimize
your Microsoft Word program by
clicking on the first of the three small boxes at the top right hand side of
your screen. The box you want is gray, square, and contains only a small line
through the box.
4. Now
your SPSS output screen should be
open in front of you.
5. Using
your mouse, click once on the Descriptives
table. The Descriptives table should
now be surrounded by a box.
6. Using
your mouse, move your pointer to the top of the SPSS output screen to the command Edit. Click on Edit.
7. Click
on Copy Objects.
8. Now
minimize your SPSS output screen by
clicking on the first of the three small boxes at the top right hand side of
your screen. The box you want is gray, square, and contains only a small line
through the box.
9. Your
Microsoft Word program should now be
in front of you.
10. Click
anywhere on the screen. There should now be a blinking cursor. Move the cursor
down several lines (in case you want to add a title or a sentence about the
SPSS descriptives table) by clicking on Enter
several times.
11. Using
your mouse, click on the Edit
command at the top of your screen. Click on Paste.
12. Your
SPSS Descriptives table should now
appear in your Microsoft Word
document. The Microsoft Word table
should be identical to the SPSS
table.
SPSS Lab Exercise 3
Frequencies
Data manipulation: Recoding and Selecting Cases.
Central Tendency Measures
Histograms
Task 1
Open your data file from Exercise 1.
Imagine that you need to classify your respondents into five
categories in terms of their ages. To do so you will need to create a new
categorical variable.
Recode the continuous
variable Respondent's Age (age) into
a new categorical variable (agegroup).
The values for the new variable will be as follows:
New Values (agegroup) Old
values (age):
1 – late adolescent 18-20
2 – young adult 21-40
3 – middle adult 41-60
4 – late adult 61-90
Note: The width
of these intervals are not equal. In a true
study, we would want the interval widths to be consistent!
In the menu bar go to Transform
·
Recode
·
Into different variable…
·
Transfer “age”
into Output variable box
·
Type the name of a new variable - agegroup
·
Click on Change
·
Click on Old and New Values
·
In Old values select Range and type the first
range of the old values: 18-20
·
In New value type 1
·
Click on Add
·
Repeat these steps for all old and new values
·
Continue
·
OK
You should have a new variable (agrgroup) with the values 1
to 4.
Now define the new
variable and its value levels. (You do this under variable view)
Now obtain the frequencies for agegroup:
What age group category has the least number of
participants/people? _________
What age group category has the most number of
participants/people? _________
What % of the sample is late adult? _________
What % of the sample is young adult? _________
What % of the sample is middle adult? _________
Task 2
Run the frequencies for the following variables: maeduc and
age.
Now find the standard deviation, variance, minimum and
maximum values for these variables.
To do so, in the main menu bar go to:
·
Analyze
·
Descriptive Statistics
·
Frequencies
·
Click on the selected variables in the left box
and transfer them to the Variable(s) box by clicking the right arrow.
·
In the same window click on Statistics…
·
Select appropriate statistics
·
Continue
·
Charts…
·
Histogram
·
Select Display normal curve
·
Continue
·
OK
|
Variable
|
Mean
|
Median
|
Mode
|
Maximum
|
Shape
|
|
Age
|
|
|
|
|
|
|
Maeeduc
|
|
|
|
|
|
Task 3
Run frequencies for the variables age and years separately
for males and females.
To do this we need to select the cases according to
respondent's gender.
To run the frequencies for each gender we will first select
males, run the frequencies for males, and then select females and run the
frequencies for females.
To select males:
·
Go to Data
·
Select cases
·
If condition is satisfied
·
If…gender =1 (Select Gender, click arrow, then
select function =1)
·
Continue (Please
note: Unselected cases should be FILTERED
as deleting the cases will delete them forever!)
·
OK
You have now selected only the males. Until you re-select
everyone, reset the select feature, or
select only females, all the statistics you do from this point forward will be
based only on males!
Next we need to run the frequencies:
·
Analyze
·
Descriptive Statistics
·
Frequencies
·
Statistics
·
Move maeduc and age variables into Variable(s)
box.
·
Click on the boxes of standard deviation,
variance, skewness, minimum, maximum, mean, mode, and median.
·
Continue
·
OK
Your SPSS output will present your frequencies for males
only. Note your sample size is smaller
then it was during task 2. We have excluded
the females from this analysis!
Next, you will have to repeat these steps for analyzing the
data for females.
Before selecting females you will need to reset the data. In
order to do so go to
·
Data
·
Select Cases
·
Reset
Now select only females using the following procedure and
then re-run the frequency analysis.
·
Go to Data
·
Select cases
·
If condition is satisfied
·
If…gender =2 (Select Gender, click arrow, then
select function =2)
·
Continue (Please
note: Unselected cases should be FILTERED
as deleting the cases will delete them forever!)
·
OK
Use your output to answer the following questions:
|
Variable
|
Mean
|
Median
|
Mode
|
St. deviation
|
Variance
|
Shape
|
|
Age/ M
|
|
|
|
|
|
|
|
Maeduc/ F
|
|
|
|
|
|
|
Now produce the histogram with normal curve for these
variables.
Remember to select cases appropriately for each variable.
In the main menu bar go to:
·
Graphs
·
Histogram
·
Transfer appropriate variable to Variable bar
·
Select Display normal curve
·
OK
Another way to obtain the frequencies for males and females
separately would have been to go to:
·
Data
o Split
File
§ Compare
Groups
·
Click on “Organize output by groups”
o Move
the gender variable from the left box into the middle box under “Groups Based
On”
§ Click
OK.
In order to unsplit the file, go back to Split
File and select “Analyze all cases, do not create groups”.
ADDITIONAL SPSS TECHNIQUES
Simple Correlation
Task 1:
Correlation between Two Variables
Use the 1991 U.S. General
Social Survey.dat data set (ITS website) to find the strength of the
relationship between fathers’ education level (highest year of school
completed, father: paeduc) and mother’s education level (highest year of school
completed, mother: maeduc).
In the main menu bar go to:
·
Analyze
·
Correlate
·
Bivariate
(meaning 2 variables)…
·
Transfer the
appropriate set of variables to the Variable box
·
The default
options selected are Pearson Correlation Coefficient, 2 tailed significance
test, flag significant correlations
·
OK
|
|
paeduc
|
maeduc
|
|
paeduc
|
r=1.00
|
|
|
maeduc
|
|
r=1.00
|
Is the correlation
significant? Yes / No If yes, at what significance level?
How many people are in the
data set?
What proportion of variance
in maeduc is explained by paeduc? ____________
Note about interpreting significant correlations: With larger samples, small correlations may be
deemed significant because of the power. A better way of interpreting
correlations is to consider the proportion of variance (r2). For
example, a correlation of 0.2 may be significant, but accounts for only 4
percent of the variance.
Scattterplot: The scatterplot
enables you to see whether a correlation will accurately summarize the
relationship between 2 variables. Correlations are appropriate only for linear
relationships. The r will be an underestimation if the relationship is curvilinear.
It is important to examine scatterplots when studying relationships between
variables.
To produce a scatterplot for
the pair of variables, in the main menu bar go to:
·
Graphs
·
Chart Builder -
OK
·
Select “Scatter”
from the gallery
·
Select “Simple”
or the first graph presented – running your mouse over each example graph will
tell you what they are.
·
Select the
variables from the list on the upper left and drag and drop the variable on the
selected axis
·
Transfer maeduc
to the Y-axis and paeduc to the X-axis
·
OK (The graph
will then be entered into your viewer folder)
SPSS produces simple
scatterplots this way. To obtain a line of best fit (more on this next lab)
·
Double click on
your graph
·
Chart Editor
window will open
·
From the menu bar
in the Chart Editor window select ELEMENTS - Fit Line at Total
·
OK
·
Close the Chart
Editor window
Describe the relationship
between the maeduc and paeduc.
Task 2:
Correlations for a Subset of the Sample
Determine the relationship
between education (educ) and mothers education (maeduc) for male students.
Reduce your output. To select
a subsample of students you need to select cases. In the main menu bar:
·
Data
·
Select Cases
·
If condition is
satisfied
·
If
·
Move Sex into
empty box on the right and create statement specifying the gender of interest
(i.e., sex = 1 will specify males)
·
Continue
·
OK
Now run the correlation
(analyze, correlate, biverate) and produce the scatterplot.
|
Male respondents
|
Education
|
Mother’s Education
|
|
Education
|
|
|
|
Mother’s Education
|
|
|
What proportion of variance
in Education is explained by Mother’s Education for male students? ____________
What do you conclude?
Before running further
analyses, you need to unselect the cases (Data, Select Cases, All Cases, OK).
Independent
Samples and Dependent Samples t-tests
Task 1: Confidence Intervals
Using
the GSS93 subset.sav data set (located within the SPSS program or on the ITS
website), you will make interval estimates (confidence intervals) of the
parameters for the adult population of the United States.
To
get the confidence intervals you will need to go to:
Ø Analyze
Ø Descriptive Statistics
Ø Explore
Ø Transfer your variable to
the Dependent List.
Ø Select Statistics:
Descriptives and specify the appropriate Confidence Interval for the Mean.
Ø Continue, OK
1. What
is the average number of years of education (mean highest level of education)
of the females adult
population?
|
|
Number of Cases
|
Mean
|
95% Confidence Interval
|
|
|
|
|
|
|
Lower Bound
|
Upper Bound
|
|
|
Years of Education (educ)
|
|
|
|
|
|
Task 2: Testing a Hypothesis About Two Related Means
Use
the Anxiety2.sav data set (ITS website).
1. Create a new variable
that is the difference between trial 1and trial 4 anxiety (variables trial1 and
trial4)
Go
to:
Ø Transform
Ø Compute
Ø Type "diff" in
target variable box
Ø Click on " trial1"
and transfer it into the numeric expression box.
Ø Click on the subtract sign
or type in "-"
Ø Click on " trial4"
and transfer it into the numeric expression box,
Ø OK
Now
make a Histogram of the variable "diff" (go to Graphs, Chart Builder,
Select Histogram, and put diff on X axis, OK) and examine the distribution.
1.
Does the distribution appear to be normal? ____________
2.
Conduct a t-test for dependent samples
Ø Analyze
Ø Compare Means
Ø Paired Samples T-test
Ø Highlight both trial1 and
trial4 variables and transfer them into the Paired variables box.
Ø Under Options specify the
confidence intervals as 95%
Ø Continue
Ø OK
3. Now answer
the following questions:
a) What is the correlation
between trial1 and trial4? _______________
b) Using the 0.05 level of
significance, do you reject or retain the null hypothesis? ___________
Task 3: Testing Hypothesis About Two Independent Means
Problem:
A researcher is interested in the effect of an approach to teaching graduate
statistics on statistics anxiety. The statistics course offered by the
Educational Psychology department is a lecture based course and a computer
based course with no lectures. The content of both courses is exactly the same.
There are twelve students in each class. At the end of the course students were
asked to fill out the Statistics Anxiety Questionnaire. The results are
presented below:
EDPY 500 EDPY
500
Lecture
Based Approach Computer
Based Approach
10 27
23 24
11 15
17 19
7 17
4 21
18 26
11
17
11
20
14
29
10 27
19
22
Please
enter this data into SPSS. (HINT: To do this, you will have to enter two rows
of data: one for the class (the first 12 rows will have an indicator 1 to
indicate lecture and the second 12 rows will have an indicator 2 to indicate
computer) and one column for the respective anxiety scores).
Test
the null hypothesis that the difference between the mean anxiety score of the
students taking the lecture based course and the mean anxiety score of the
students taking the computer based course is zero.
1.
Enter the data into the SPSS file and define the variables.
2.
Produce the histograms and examine the distribution of the anxiety scores for
both groups.
To
do this go to:
Ø Data
Ø Split File
Ø Click on "Organize
Output by Groups"
Ø Click on Groups Based On:
Ø Enter Class
Ø Sort File By Grouping
Variable
Ø OK
3. Do the scores in both populations appear to
be normally distributed?
4. Go Back and UNSPLIT the file. Remove
"class" from Groups Based On, Click on Analyze All Cases
and then select OK.
5. Conduct a t-test for two independent samples:
Ø Analyze
Ø Compare Means
Ø Independent Samples t test
Ø Transfer your dependent
variable (anxiety) to Test Variable(s) and the independent variable (teaching
approach) to the Grouping Variable bar.
Ø Define the groups...
Ø Type the numerical values
for the two groups
Ø Continue
Ø Under options select the 95%
confidence interval
Ø Continue
Ø OK
6.
Examine your output and answer the following questions:
a)
What are the mean anxiety scores for the two groups? _____________ _____________
b)
Is the assumption of homogeneity of variance met? For Levene's test for
equality of variances, if
the test is nonsignificant, do not reject
the hypothesis that the two population variances are equal. _________
c)
What is the mean difference for the two samples? ______________
d)
What is the value of the t test? ______________
e)
How many degrees of freedom are there? ______________
f)
What is the obtained p value? ______________
g)
Using the 0.05 level of significance, do you reject or retain the null
hypothesis? ______________
Comments