lectures (topic-2, lec-4)
TRANSCRIPT
-
8/7/2019 Lectures (Topic-2, Lec-4)
1/5
Training Course on SPSS for Professionals, Researchers an
Students
Lecture 04: Data Manipulation II
Data Manipulation II includes:
A. Splitting File
B. Case Selection
C. Selecting a Random Sample
D. Weight Cases
To learn the above manipulation techniques, let us enter the following data into SPSS and save it in
computer.
Subject Sex Age Religion EducationIncome per
day
1 1 25 1 15 100
2 1 32 2 15 200
3 1 30 3 16 150
4 2 24 1 13 125
5 2 20 2 12 1206 2 27 3 19 125
7 1 25 1 12 175
8 1 33 2 13 200
9 2 39 3 15 120
10 2 40 1 17 125
11 1 32 2 11 135
12 2 27 3 10 130
13 1 37 3 14 100
14 1 19 2 12 150
15 2 25 1 15 120
Sex : 1 = Male, 2 = Female
Religion : 1 = Muslim, 2 = Hindu, 3 = Christian
A. Splitting File
You may encounter some situations where you need to analyze data based on the categories of one
or more categorical variables. For example, you might seek to know the picture of income or
education among the sexes either for comparison purposes or for organizing the output by groups.
14th Batch (March April, 2011)Page 1 of 5
-
8/7/2019 Lectures (Topic-2, Lec-4)
2/5
Training Course on SPSS for Professionals, Researchers an
Students
To do this you have the SPSS tool Split File.
Split File splits the data file into separate groups for analysis based on the values of one or more
grouping variables. If you select multiple grouping variables, cases are grouped by each variable
within categories of the prior variable on the Groups based on list. For example, if you select Sex as
the first grouping variable and Religion as the second grouping variable, cases will be grouped by
Religion classification within each Sex category.
Note that you can specify up to eight grouping variables and cases should be sorted by values of the
grouping variables.
1st Approach: Group Comparison
Suppose you want to compare income of the sexes. Then it would be convenient if the results are
shown in a single table for both sexes separately. So to get output in that format you should do
splitting the data file. And you can split your data file just following the instructions given below:
Click on Data (Menu Bar) and then Split File
Select Compare Groups
Now select the variable Sex and send it to the box Groups Based on:
ClickOK
Now if you carry out any analysis on this data file you will get the results in your desired format.
Again you may seek to compare income among people of different religions and sexes at a time.
Repeat the above steps. Here with Sex variable select the variable Religion and send it to the box
Groups Based on, click OK.
2nd Approach: Organizing Output by Groups
Suppose you want the results are displayed separately in different table for each split-file group.
For example, you might desire to summarize the information on education variable for people of
varied religions separately in different tables. To do this, follow the instructions given below:
Click on Data (Menu Bar) and then Split File
Select Organize output by groups
Now select the variable Religion and send it to the box Groups Based on:
ClickOK.
14th Batch (March April, 2011)Page 2 of 5
-
8/7/2019 Lectures (Topic-2, Lec-4)
3/5
Training Course on SPSS for Professionals, Researchers an
Students
Similarly try for other variables (say Sex).
You can use the Toolbar item to Split a data file. Just click on the Split File tool. And the Split File
box will appear. Now you can do all the steps as before.
B. Case Selection
Sometimes you may have special interest on particular cases only. To cope with this situation,
SPSS will have for you the Select Cases option. For instance, suppose you want to analyze data
considering only male respondents/cases. Then you have to select only male cases. To do it, use the
following instructions:
Click on Data (Menu Bar)
Select Cases
Select If Condition is Satisfied
Click on If
Select the variable sex and send it to the expression box.
Click on = sign on the calculator pad, click on 1 and then Continue
Select Unselected Cases are Filtered
Click on OK.
Again suppose you want to study only Hindu and Christian people. Then select these cases
following the instruction given below:
Click on Data, Select Cases, Select If condition is satisfied
Click on If, send the variable Religion to the expression box
Click on sign and 2 in the calculator pad
Click on OK.
Again you might want to study only Muslim females. You can select the desired cases simply by
taking Sex and Religion variables in the expression box of Select cases: If and writing them as
follows:
sex = 2 & religion = 1
Now click on Continue and then OK.
14th Batch (March April, 2011)Page 3 of 5
-
8/7/2019 Lectures (Topic-2, Lec-4)
4/5
Training Course on SPSS for Professionals, Researchers an
Students
C. Selecting a Random Sample
You can also take random sample of cases. The sample size may be an approximate percentage of
all cases or you can exactly mention the sample size. Suppose you will select a random sample of
size 40% of all cases in your data file. To do this, follow the instructions below:
Click on Data (Menu Bar)
Select Random Sample of Cases
Click on Sample
Select Approximately and write 40% in the boxClick on Continue and then OK.
Again suppose you will select a random sample of size 7 from all cases in your data file. To do this,
follow the instructions below:
Click on Data (Menu Bar)
Select Random sample of Cases
Click on Sample
Select Exactly and write 7 in the 1st box and 15 in the 2nd box
Click on Continue and then OK.
You can also take a sample just mentioning the range of cases. For example, you may need to study
some consecutive cases from anywhere of the data file (say, cases from 5 to 12). To do this, follow
the instructions given below:
Click on Data (Menu Bar)
Select Based on time or case range
Click on Range, write 5 in the First Case box and write 12 in the Last Case box
Click on Continue and then OK.
D. Weight Cases
Weight Cases gives cases different weights for statistical analysis. The values of the weighting
variable should indicate the number of observations represented by single cases in your data file.
Note that Cases with zero, negative, or missing values for the weighting variable are excluded from
14th Batch (March April, 2011)Page 4 of 5
-
8/7/2019 Lectures (Topic-2, Lec-4)
5/5