SPSS Training Report
SPSS Training Report
SPSS Training Report
On
SPSS Training
Statistical Package for Social Sciences
FATA Research Centre holds a series of guest lectures on significant issues of FATA in
collaboration with The Royal Danish Embassy in Pakistan.
The purpose of guest lectures is to enhance the community base knowledge regarding
FATA. Training and workshops on capacity building is also part of the lectures. It was
crucial to train the employees of FRC to achieve this aim with high precision. In the same
connection, a guest lecture was delivered by Management Xtrategizers on SPSS.
FRC is privileged to thank the honorable guest speaker Professor Rana Qudoos, who
skillfully delivered lecture on SPSS during the workshop and to the Management
Xtrategizers that provided us such a resource person.
FRC also would like to thank its employees who participated in the training with great
enthusiasm and attention.
The report is drafted by Engr. Taimur Ali khan (IT Engineer), composed by Miss Rubab
Zahra (Research Associate), assisted by Mr. Asad kamal (Researcher), reviewed by Mr.
Mehran Ali Khan Wazir (Programme Manager) and approved by Dr. Ashraf Ali (President
FRC).
Acknowledgement.....……………….………………………………………...……………………………………. i
Table of Contents …………………………………………………………………………………………………… ii
Introduction ..........................................................................................................................................................2
Statistical Package for the Social Sciences (SPSS) ..............................................................................2
Using SPSS for the first Time ........................................................................................................................3
1. SPSS Views/Windows: ......................................................................................................................3
2. Defining Variable: ...............................................................................................................................4
3. Data Entry to the software ................................................................................................................7
Analysis..................................................................................................................................................................8
1. Introduction to Descriptive Statistics:...........................................................................................8
2. Frequency Command..........................................................................................................................9
3. Descriptive command ...................................................................................................................... 11
4. Cross TAB Command ..................................................................................................................... 12
Introduction to illustrative statistics .......................................................................................................... 14
1. Graph of Single variables (simple Chart) ................................................................................. 15
a. Pie Graph ............................................................................................................................................ 15
b. Bar chart.............................................................................................................................................. 16
2. Graph of Multi variable (Cluster Chart).................................................................................... 16
List of Participants .......................................................................................................................................... 22
Agenda ................................................................................................................................................................ 23
Being a research based independent and non-political institute, FATA research centre
attaches immense importance and gives utmost concentration in presenting its products in a
more professional way. With this aim, FRC conducted a training session on SPSS for its
employees to not only enhance their skills but also to able the product presentable on the
national and international forums. These sort trainings are the part of FRC series of capacity
building sessions for its employees and for other eligible professional from FATA.
SPSS is a great tool of data handling and analysis in social sciences research. Other tools
such as excel and other database are also used for the same purpose but SPSS provides an
easy access to descriptive statistics and frequencies (Responses), that gives a wider variety
of charts & graphs, better and more flexible pivot tables and provides a full set of statistical
tests. Using the “Split” function in SPSS, it is relatively easy to create tables and charts for
subsets without doing any extra work. An attractive feature of SPSS is Labeling Data instead
of codes in reports, because the survey software that converts variable Male- into 1 and
variable Female- into 2 doesn’t mean you want to see lots of 1s and 2s in the report. SPSS
provides an accurate result even when some data is missing, while it also provides easy
import functions, unlimited rows for entries and many additional features. In short SPSS
saves time and increases productivity and makes it easy to understand statistical results.
SPSS installation is quite an easy job. Just double click on Software setup and follow the
instructions. After installation one should access SPSS by double clicking its desktop icon.
A window opens, before data editor screen, shown in Figure - 1, gives one an opportunity to
run a self-help tutorial, type in data, run an existing query, create a new query using the data
base wizard, open an existing data source or open another type of file ( file created in MS
Excel). For first time users, one should obviously go for the 2nd option, “Type in Data”.
1. SPSS Views/Windows:
Choosing one of the above mentioned options would go through three types of different
“views”, “Data view”, “Variable View” and “Output View.” One can see the toggle switch
for the first two views at the bottom left hand corner of Figure – 2, while the third view is
used for observing DATA results.
Figure 1 Figure 2
Once you install and open SPSS you are now in the position to design your survey. By
designing we mean the definition of variables for your survey and set properties (Attributes)
related to each variable. Variable refer to the survey questions (Age, Gender, Marital
Status, Income etc).There are two main reasons to fully define a variable:
To keep an electronic archive of your data and all its attributes (e.g. complete survey
items; coding scheme)
To set up your data so that when you start to run analyses, the results and any tables
and graphs you produce will be attractive, complete, and understandable to the
reader.
To start defining variables one should go to variable view by clicking the variable view tab.
In that view sheets in the column headings have now changed from ‘var’ to the ones shown
in Figure – 3:
Figure 3
The sheet is no longer the data entry sheet the user originally was working in; it will now be
set up differently, because now you are in Variable View. Many new users of SPSS get
confused when switching between Data view and variable view sheets. This is because in
data view the variables are in columns so the data should be entered and read moving down
the sheet while in variable view the variable are in rows. Therefore, the variable information
should be entered and read moving across the sheet.
To set upon a scenario and consider how each variable is initialized we will go through the
FRC Research Study as an example. The purpose of the research was to closely
observe/monitor the developments taking place in Afghanistan after post withdrawal of
US/NATO forces, and their effect on FATA. Sixteen different questions including seven
demographic questions (Agency, Age, Gender, Marital Status, Education Religious,
Education Non-Religious and Occupation) are asked from the respondent. A survey of more
than 2000 questioners is done and then analyze in SPSS. It is now clear that our design
should compose of 24 Variables. One extra variable is representing Data Entry Operator
Information. To set up each variable one should set all the attributes of a variable in variable
view according to their design. All attributes are explained below:
Name refers to “variable name”, in row 1, 2, and 3; and so on below the Name column in
Variable view enter the name of your variable. In the FRC research study the variables are
DE Operator, Agencies, Age, Gender, and Marital Status and so on.The first character of
each variable must be a letter. Name is like the shortest identity of a variable(question) i-e
STRING variables include letters, and can be used for names or perhaps brief open-
ended responses
NUMERIC variables are numbers. In either case, you’ll need to specify your
variable width. You can do that in the dialogue box, or in the subsequent WIDTH
and DECIMAL columns
Since variable names often tend to be very short LABEL allows one to specify a longer
variable name. For question six of FRC Research Study you can give the full question string.
Variable label play very important role, and will appear on charts or graphs during analysis.
“Data Entered by an Operator” can be a label for variable “DE Operator.”
Value labels provide a method for mapping your variable values to a string label. In this
example, there are three acceptable values for the M-Status variable. A value of 1 means that
the subject is single and a value of 2 means that he or she is married while 3 shows that the
subject has a different marital status . This helps us in the data entry process because it’s a
tough practice to enter strings instead of integers. When the subject clicks the cell and then
clicks on the button with the three dots, a dialogue box will open.
Figure 4
For male write 1 in
value text box and
Male in label
textbox and press
Add button.
Columns give width to the columns in data view. Normally this would be at least 8 so that
the variable name could appear easily but one can change accordingly. The subsequent
column, Align, allows you to change the presentation so that scores for the variable are left-
justified, right justified, or centered.
Measure refers, for numeric variables, to the level of measurement of one’s variable. It
distinguishes three variables: Nominal, Ordinal and “Scale”.
Nominal: A variable can be treated as nominal when its values represent categories
with no intrinsic ranking (for example, the department of the company in which an
employee works). Examples of nominal variables include region, zip code, and
religious affiliation.
Ordinal: A variable can be treated as ordinal when its values represent categories
with some intrinsic ranking (for example, levels of service satisfaction from highly
dissatisfied to highly satisfy). Examples of ordinal variables include attitude scores
representing degree of satisfaction or confidence and preference rating scores.
Scale: A variable can be treated as scale when its values represent ordered
categories with a meaningful metric, so that distance comparisons between values
are appropriate. Examples of scale variables include age in years and income in
thousands of dollars.
The overall study design of FRC is represented below in Figure – 6. A total of 25 variables
are created to enter data, and analyze it. First variable DE Operator gives the information of
how many questionnaires have been filled by each data entry operator. Agencies variable
holds information regarding to which agency the respondent belongs to. Six demographic
questions, (Age, Gender, M-Status, Edu-Relg, Edu-Non-Relg, and Occupation) records
respondent personal information.
Figure 6
Once all variables are defined completely, it is time to start entering data. As discussed
earlier in the variable definition that a special code is to be assigned to each variable in a
questionnaire. These codes will now be entered for each and every variable in a ll variables
for the questionnaire and so on for others. The data view columns shows that how the
variable (All Question from a Questionnaire) and each row shows the case/respondent.
Figure 7
Analysis
Analysis can be made in different ways like descriptive statistics, illustrative statistics,
different statistical test etc. this report present descriptive statistics and illustrative statics.
Descriptive statistics are those that describe the relationship between variables(like finding
the ratio of male and female in a Survey, find their response about a question, accurately
describing large datasets quickly and easily.).The most common descriptive statistics used
are the measures of central tendency (mean, median and mode) and the measures of
dispersion (standard deviation, standard error and variance).
The commands for extracting descriptive statistics come under the “Analyze/Descriptive
statistics” dropdown as show in figure - 8. Various command come under the dropdown,
such as” Frequency”, “Descriptive”, “Explore”,and Crosstabs etc
Figure 8
The Frequency command allows you to analyze a full range of descriptive statistics
including the measures of central tendency, percentile values, dispersion and distribution.
From the Analyze drop-down menu, select Descriptive Statistics and then Frequencies.
Figure 9
Select Gender or any of the
variables, we are taking Gender as
an example and click on the arrow
button between the variables list and
the Variable(s) box to send it across.
Table 1
S tatistics Where, N tells about the Valid and Missing values
in the Gender column. Clear from the table, FRC 3rd
Gender
study is composed of 2051 respondents, Missing are
N Valid 2051 zero because missing data forms are discarded from
computation.
M issing 0
The Table– 2 with the heading Gender gives a frequency breakdown for males and females:
Table 2
Gender
The Frequency column shows a frequency count for each gender (male and female) and
also includes the total count for both. The table also shows percent, valid percent, and
cumulative percent. The different types of percentage are very helpful in analysis they show
clear proportions of each case in a variable. One can find the frequency and percentage for
each and every variable in study.
Select the variables you wish to include in the descriptive analysis. We will again
use the FRC study as an example. Send any of the variable to the Variable(s) box
Table 3
Statistics
M issing 0 0
M ode 2 4
M inimum 1 1
M aximum 3 6
S um 3654 5355
If one wishes to support their descriptive statistics with some charts. The Charts button is
very useful for producing more than one chart at a time, including bar charts, pie charts and
histograms. Histograms can help identify if your variables are normally distributed. In our
example we want to explore the distribution in Question 7 and question 13:
Click on the Charts button. Select the Histograms radio button and also click in the
box for with normal curve
Click Continue
Deselect the Display frequency tables and All the Previous Checked Statistics on
this occasion as we only want the charts and then click OK
Figure 11
The above mentioned graphs show the histogram including Normal curve, that give some
useful information regarding question 7th and question 13th .
3. Descriptive command
The following table (Table - 4) shows the Descriptive Statistics for the variables
‘Question_7’ and ‘Question_13’:
Table 4
Descriptive Statistics
Minimu S td.
N m Maximum Mean Deviation
The Descriptive Statistics table consists of the number of participants (N) and the
Minimum and Maximum Response for each Question. It also shows the Mean and the
Standard Deviation for each variable .
Crosstabs is also a useful feature of SPSS; they are used to analyze demographic
information. By demographic information we mean to find the frequencies of dependent
variable (Row Variable) over independent variable (Column Variable). For example we are
taking ‘Age’ as independent variable and ‘Question_7’ as dependent variable.
click on ok
Table 5
Age * The individuals or Quam will respond aggressively to the militancy in the post withdrawal scenario.
Cross-tabulation
% within
41.8% 46.7% 11.5% 100.0%
Age
% within
37.5% 46.9% 15.5% 100.0%
Age
% within
35.0% 51.2% 13.8% 100.0%
Age
% within
27.0% 57.8% 15.1% 100.0%
Age
% within
36.1% 49.6% 14.3% 100.0%
Age
Figure 14
From the above graph we can also analyze that for all the cases the response of disagree is
more than the other two. Beside cross tabulation one can draw different types of table from
Analyze/tables/custom tables… dropdown menu.
Illustrative statistics are a visual representation of data. They can be useful tools for
summarizing and comparing some or all data. Illustrative statistics is an easy way to study
general trends of data in graphs rather than trying to interpret numerical information. There
are three ways of obtaining illustrative statistics in SPSS.
1. All types of graphs can be obtained using the Graphs drop-down menu.
2. You are also able to produce some graphs using the Frequencies command as
shown earlier in the report
3. Finally, many statistical test procedures have chart options. But they are beyond the
scope of this report. One can find many tutorials available online and in hard copy
on these.
From the Graphs dropdown select Chart Builder option, the following Window will
appear:
In the bottom left box one can Select chart type (Bar, Line, Area, Histogram) while clicking
on a type one can also choose sub type. After selecting your desired type just click and drag
the type to the white upper right window.
Figure 16
Figure 17
Chart showing data for more than one variable is also available in SPSS. Suppose one of the
researchers at FRC want to analyze Agency wise and Frontier Region wise results for a
variable. Simply go to the chart builder select and drag multiple lines to the white window. I
have used this type of graph out of personal preferences there are several other options (Bar,
Histogram, etc...)
Figure 18
One can do a number of operations to their data, to make new variables from the current
variables, sort the existing variables, apply transpose to the existing variables, to transform
their data for different purpose, to select a portion of data for analysis, split and merge data
files etc.
These commands are in the Transpose and Data dropdown. Some of these commands are
discussed for other commands one can visit FRC and References links.
Compute Variables:
The Transform drop-down menu offers a variety of commands which allow you to
transform your data, such as adding scores together or recoding data into a new variable.
Compute command does exactly what its name says; it enables the user to perform
arithmetical operations that usually result in a new variable being produced.
For example, you may have collected data from a questionnaire and need to add make a
variable representing the missing values of a variable, this is possible within the Compute
procedure.
A lot can be done in this window, a lot of built in functions are presented in SPSS one can
create new variable of their own choice. The procedure here for selecting a specific function
is to first selecting a function (parent function, lies in “Function group:”) then choose sub
function from “function and special variables” and drag it to numeric expression box. I-e
FRC wants to create a new variable that measure Missing values in a variable. Parent
function case example for finding missing values in a variable is missing values and Sub-
type Nmiss. By double clicking on Nmiss, expression would appear in Numeric Expression
window. NMISS works with multiple numeric values, whereas MISSING works with only
one value that can be either numeric or character.
Figure 21
Figure 22
Select Case:
Questioners have been filled by respondents of 7 agencies and 6 FR Regions. Suppose FRC
is interested in analyzing individual Agencies or FR Regions, and comparing that to the
collective agencies response. Select Case command will help them to analyze a specific
case. In FRC Study the Agencies Variable are label form 1=SWA, 2=NWA, 3=KURRUM
… 13=FR Peshawar. Through select case command graphs for SWA and NWA are
computed and compared with the combined FATA Graph (Figures – 23 and 24)
The above mentioned graphs can be created via Select case as shown in Figure - 26
Figure 27
List of Participants
Guest Speakers:
1. Professor Dr. Rana Qudoos.