INTRODUCTION OF SPSS 15.0 FOR WINDOWSLab Activity Community Research Program IV STUDENT’S GUIDE INTRODUCTION OF SPSS 15.0 FOR WINDOWS Faculty of Medicine Universitas Padjadjaran

Lab Activity Community Research Program IV

STUDENT’S GUIDE

INTRODUCTION OF SPSS 15.0 FOR WINDOWS

Faculty of Medicine

Universitas Padjadjaran

JATINANGOR

2012


Week:

Introduction of SPSS (two meetings)

Learning Objective:

At the end of the course, the student should be able to:

1. Start the SPSS 15.0 program

2. Perform data entry

3. Perform data import from excel file (xls)

4. Save data in SPSS file (sav) and SPSS output file (spo)

5. Perform data sorting and selection

6. Perform data transformation as follow:

a. Involving dates and times

b. Recoding variables

7. Perform descriptive statistics as follow:

a. Frequency and percentage

b. Mean, median

c. Standard deviation, range, quartiles, interquartile range

8. Create an appropriate table, chart, or graph using SPSS to present the data

9. Open an existing data in SPSS file or SPSS output

Reference

1. Peat J, Barton B. Medical Statistics: A Guide to Data Analysis and Critical Appraisal. 1st ed.

Oxford: BMJ Books; 2005.


1. Starting the SPSS 15.0 for Windows program

Click Start menu

Click All Programs

Choose SPSS for Windows

or

Click the shorcut

You will have the following window:


Select Type in data

Click OK

Now, we have two windows, run in different settings. The first is run under SPSS Data Editor

(untitled1 dataset0) and the second is run under SPSS Viewer (output1 document1).

The first window consists of Data View and Variable View. In Data View, the appearance is

almost similar with Microsoft Office Excel, having rows and columns. The difference is that we

are able to modify the columns by giving the variables name, in Variable View, thus to enter the

data, we can start from the first row. In Excel, to enter the data, we usually start from the second

row, as the first is being use for variables name.


2. Performing data entry

2.1 Variable setting in Variable View

2.1.1 Name

Must begin with an alphabetic character and each must be unique. To use space

between words is not allowed, instead we can use underscore ( _ ).

2.1.2 Type

In this occasion, we are going to use Numeric and Date.

Click

Click


2.1.3 Width

For variable type numeric, it shows how many numbers/digits, thus it can be

modified in previous step.

2.1.4 Decimals

For variable type numeric, it shows how many numbers/digits behind coma, thus it

can be modified in previous step.

2.1.5 Label

Further description about the variable can be type here and will be shown on the

SPSS output replacing the variable name, after performing a statistical analysis

(descriptive or analytic) or creating chart/graph. If it were left empty, then the

variable name would be shown instead.

2.1.6 Values

When we use variable type numeric to represent categories (let‟s say for variable

sex, we assign 1 = male and 2 = female; for variable socioeconomic status, we assign

0 = high, 1 = moderate, and 2 = low) or after recoding the variables (further

discussion in step 6.2), make sure we put value label for each number. The order of

the number is important as it will determine the data presentation after performing a

statistical analysis (descriptive or analytic) or creating chart/graph. For instance, if

we use a table, the smallest number will be positioned on the first row or on first left

column.


If we do not put any value label, we will have unappropriate output, since it only

shows the number and whoever read the output will be confused or have wrong

conclusions, especially for those unfamiliar with the data or the study. As shown in

the dummy table, one might have a conclusion: from the study, we have 11 persons

with low, 11 persons with moderate, and 8 persons with high socioeconomic status.

What is the conclusion from following dummy table, which value labels have been

assigned correctly?

2.1.7 Missing

Suppose we might want to distinguish between data that are missing because a

respondent refused to answer and data that are missing because the question didn't

apply to that respondent. Data values that are specified as user-missing are flagged

for special treatment and are excluded from most calculations. In this occasion, we

will have a complete set of data, thus the default setting will be no missing values.

2.1.8 Columns

If necessary, we can set the width of the column in the Data View.

2.1.9 Align

If necessary, we can set the alignment in the Data View.


2.1.10 Measure

Nominal scales:

1) Have no order and are generally category labels that have been assigned to

classify items or information

2) In SPSS, it can be string (alphanumeric) values or numeric values that have been

assigned to represent categories

Ordinal scales:

1) Have a logical or ordered relationship across the values

2) It is usually not possible to measure a specific amount difference between

categories

3) In SPSS, it can be string (alphanumeric) values or numeric values that have been

assigned to represent categories

Scale

1) In SPSS, set to this measure if the variable type numeric (interval or ratio) or date

2) In SPSS, set to this measure for numeric values that have been assigned to

represent categories (do not forget to put value labels!)

2.2 Data Organization

2.3 Enter the data in Data View

To practice data entry, please use Data entry exercise for SPSS.pdf (can be downloaded

from crpfkunpad.multiply.com). After the information for each variable has been defined in

Variable View, the data can be entered in the Data View screen. Try to perform data entry in

10 – 15 minutes!

3. Performing data import from Excel file (xls)

Many researchers use Excel or Access for ease of entering and managing the data but for

statistical analyses and creating appropriate charts/graphs are best executed in a specialist

statistical package as SPSS, as the integrity and accuracy of the statistics are guaranteed. As for

SPSS 15.0, it can import data in an Excel 2003 (xls) and Access 2002 -2003 (mdb) file format.

Starting from SPSS16.0 version, it can import until Excel 2007 (xlsx) and Access 2007 (accdb)

file format. Make sure we do not open the file using MS Excel while we try to import the data

into SPSS!


Suppose we will import Measles Case data in Surveillance Data Analyses 2012.xls file into

SPSS. We can use the following steps:

3.1 Open Data Method

Click File

Select Open

Select Data…

or by Clicking symbol under File tab

On Open Data Window:

Search the intended folder

Select Excel (*.xls) in Files of Type options

Select Surveillance Data Analyses 2012.xls file

Click Open

On Opening Excel Data Source window:

Put a tick/check mark on Read variable names from the first row of data option

Select Measles Case in Worksheet options

Change the maximum width for string columns as 15 or 20 (not to wide)

Click OK



3.2 Open Database Method

Click File

Select Open Database

Select New Query…

Select Excel File

Click Next

or double-click on Excel File

On the ODBC Driver Login window:

Click Browse…

On Open Data window:

Search the intended folder

Select Excel (*.xls) in Files of Type options

Select Surveillance Data Analyses 2012.xls file

Click Open


On the ODBC Driver Login window:

Click OK


On Database Wizard window:

Just follow the instruction to select the data. An alternative choice is by double-clicking

„Measles Cases$‟

Click Next

Click Next again when we were asked to limit the cases

If we allow SPSS to automatically recode string variables (remember the value label we

have discussed earlier), we can put a tick/check mark for string variables. One might say this

is an advantage, but other might say no, since the value label are ordered according to

alphabetical order of the string value (Example: 1 for No and 2 for Yes). If we choose not to

recode to numeric, then we shall do it later (in step 6.2.1)

Change the Width for variable-width string fields as 10 (not to wide)

Click Next

Select Retrieve the data I have selected

Click Finish




4. Saving data in SPSS file (sav) and SPSS output file (spo)

4.1 Saving data in SPSS file (Make sure we are on Data View or Variable View screen)

Click File

Select Save

In Save Data As window:

Select the intended folder

Type “measles” as the file name

Select SPSS(*.sav) in Save as type

Click Save


4.2 Saving SPSS output (Make sure we are on Output View)

Click File

Select Save

In Save Output As window:


Type “measles” as the file name

Select Viewer Files (*.spo) in Save as type

Click Save


4.3 Save the other data from previous data entry procedure as SPSS file and name it as “data

entry spss”, so we will have two SPSS files and one output file.

4.4 Notes

Any SPSS 15.0 data file (*.sav) can be opened by different version of SPSS, let‟s say SPSS

16.0, 17.0, or 18.0, but the SPSS 15.0 output file (*.spo) can‟t be opened by different

version of SPSS, since each version have its own output file.

5. Performing data sorting and selection (please use data entry spss.sav file!)

5.1 Data sorting

Suppose we will sort the data according to date of birth of the mother (ascending)

5.1.1 By right-clicking the variable name

5.1.2 By clicking Data tab

Click Data

Select Sort Cases…

Right click


In Sort Cases window:

Select dob variable then move it to Sort by window

Select Ascending in Sort Order options

Click OK

5.2 Data selection (please use data entry spss.sav file!)

Suppose we will select the data only for smoking mothers, thus further data transformation

(step 6), descriptive statistics (step 7), and presenting data (step 8) will be applied only for

smoking mothers.

Click Data

Select Cases…


In Select Cases window:

Select If condition is satisfied

Click If…

Select

Click


In Select Cases: If window:

Select smoking variable then make sure we have expression: “smoking = 1”


Click OK


To select all data again:

Click Data

Select Cases…


Click Reset

Click OK


6. Performing data transformation

6.1 Data involving dates and times (please use data entry spss.sav file!)

6.1.1 Determining age (years) from date of birth of mother

Click Transfom

Select Date and Time Wizard…

In Data and Time Wizard window:

Select Calculate with dates and times

Click Next


Select Calculate the number of time units between two dates

Click Next

In Date 1: select Current date and time (March, 12th

, 2012) from Variables box

In minus Date2 : select dob from Variables box

Set the unit as Years

Click Next


in Result Variable, please type “age”

Click Finish


6.1.2 Predicting duedate (next 280 days from the first day of last menstrual period)

Click Transfom

Select Date and Time Wizard…

In Data and Time Wizard window:

Select Calculate with dates and times

Click Next


Select Add or subtract a duration from a date

Click Next

In Date: select first day of last menstrual period from Variables box

In Duration Constant type 280

In Unit: select Days

Click Next


In Result Variable, please type “duedate”

Click Finish


6.2 Calculating/computing (please use data entry spss.sav file!)

Determining body mass index (BMI, kg/m2)\

Click Transform

Select Compute…

In Compute Variable window:

In Target Variable: type “bmi”

In Numeric Expression: select weight and height from the left box, make sure we have

expression: weight / (height / 100) ** 2

Click OK


6.3 Recoding variables

6.3.1 Into Different Variable (please use data entry spss.sav file!)

6.3.1.1 Age group of the mother

20 – 24 years (numeric code = 1)



Click Transfrom

Select Recode into Different Variables…

In Recode into Different Variable window:

Select variable age and move it to the right box

In Output Variable, for Name: please type “agegroup”

Click Change

Click Old and New Values…


In Recode into Different Variable: Old and New Values window:

Select Range, then type 20 trough 24

In New Value, select Value, then type 1

Click Add




Click Add



Click Add

Click Continue

Click OK


Click Variable View

Modify Values for agegroup variable

In Value Labels window:

Type value 1, type Label “20 – 24 years”



Click OK

Click


6.3.1.2 BMI category/classification

18.5 kg/m2 = Underweight (numeric code = 1)

18.5 – 24.99 kg/m2 = Normal (numeric code = 2)

≥ 25 kg/m2 = Overweight (numeric code = 3)

Click Transfrom

Select Recode into Different Variables…

In Recode into Different Variable window:

Click Reset

Select variable bmi and move it to the right box

In Output Variable, for Name: please type “bmicat”

Click Change



In Recode into Different Variable: Old and New Values window:

Select Range, LOWEST through value: then type 18.49


Click Add

Select Range, then type 18.5 through 24.99


Click Add

Select Range, value through HIGHEST: then type 25


Click Add


Click Continue

Click OK


Click Variable View

Modify Values for bmicat variable


Type value 1, type Label “Underweight”

Type value 2, type Label “Normal”

Type value 3, type Label “Overweight”

Click OK

Click


6.3.2 Into Same Variable (please use measles.sav file!)

Yes = 1 and No = 2

Click Transfrom

Select Recode into Same Variables…

In Recode into Same Variable window:

Select all variables using Yes or No answer and move them into String Variables

box


In Recode into Same Variables: Old and New Values window:

In Old Value, type “Yes”, then type New Value as 1

Click Add


In Old Value, type “No”, then type New Value as 2

Click Add

Click Continue


Click OK

Click Variable View

Modify Type (from String to Numeric) and Values for these variables: Fever, fever

with rash, conjunctivitis, coryza, cough, lymphnode, probable case, confirm case,

traveling history, hospitalized, death, and immunization


Type value 1, type Label “Yes”

Type value 2, type Label “No”

Click OK

Click


7. Performing descriptive statistics (please use data entry spss.sav file!)

7.1 Frequency and percentage (categorical data)

7.1.1 Suppose we wish to know frequency of age group and BMI category for all

participants

Click Analyze

Select Descriptive Statistics

Select Frequencies…

In Frequencies window:

Select agegroup and bmicat variables and move them into Variable(s) box

Put a tick/chek mark onto Display frequency tables option

Click OK

Check the result in SPSS output



7.1.2 Suppose we wish to know frequency of age group and BMI category based on

smoking status

Click Analyze


Select Crosstabs…

In Crosstabs window:

Select agegroup and bmicat variables and move them into Row(s) box

Select smoking variable and move it into Column(s) box

Click Cells…

In Crosstabs: Cell Display

Put a tick/ check mark for Column percentages (depend on the study design)

Notes: Cross-sectional and cohort, select Row; Case-control, select Column

Click Continue


Click OK



7.2 Mean, median, standard deviation, range, quartiles, interquartile range (numerical data)

7.2.1 Suppose we wish to know mean, median, SD, range, quartiles, and IQR of age

(years) and BMI (kg/m2) for all participants

Click Analyze


Select Explore…


In Explore window:

Select age and bmi variables and move them into Dependent List box

Click Statistics…

In Explore: Statistics

Put a tick/check mark for Percentiles

Click Continue

Click OK




7.2.2 Based on smoking status

Click Analyze


Select Explore…

In Explore window:

Select age and bmi variables and move them into Dependent List box

Select smoking variable and move it into Factor List box

Click Statistics…


In Explore: Statistics

Put a tick/check mark for Percentiles

Click Continue

Click OK




8. Creating an appropriate table, chart, or graph using SPSS to present the data

8.1 Contingency table (please use measles.sav file!)

Suppose we wish to know whether sex is a risk factor for measles disease (confirm case)

using a case-control study. We will use the same method with step 7.1.2 and add Statistics

Click Statistics…

In Crosstab: Statistics window:

Put a tick/check mark for Risk

Click Continue


Click OK



8.2 Pie chart (please use data entry spss.sav file!)

Suppose we wish to know the proportion of smoking among mothers

Click Graphs

Select Legacy Dialogs

Select Pie…


In Pie Charts window:

Select Summaries for groups of cases

Click Define

In Define Pie: Summaries for Groups of Cases window:

Select smoking variable and move it into Define Slices by box

Click OK



8.3 Bar (please use data entry spss.sav file!)

Suppose we wish to know the agegroup distribution of the mothers

Click Graphs


Select Bar…

In Bar Charts window:

Select Simple


Click Define


In Define Simple Bar: Summaries for Groups of Cases window:

Select agegroup variable and move it into Category Axis box

Click OK



8.4 Histogram (please use data entry spss.sav file!)

Suppose we wish to know the distribution of BMI (kg/m2) of mothers

Click Graphs


Select Histogram…

In Histogram window:

Select bmi variable and move it into Variable box

Click OK



8.5 Line (please use the data of Clinical Scenario B from Surveillance Analyses Data Exercise!)

Suppose we wish to know the rotavirus infection trend from 1974 to 1989 in one large

district in Bandung

Click Graphs


Select Line…

In Line Charts window:

Select Simple


Click Define


In Define Simple Line: Summaries for Groups of Cases window:

Select Other statistics for Line Represents option

Select rotavirus_cases and move it into Variable box

Select year variable and move it into Category Axis box

Click Change Statistics…


In Statistics window:

Select Sum of Values

Click Continue

Click OK



8.6 Box-plot (please use data entry spss.sav file!)

Suppose we wish to compare between median of BMI (kg/m2) of smoking and non-smoking

mothers

Click Graphs


Select Boxplot…


In Boxplot window:

Select Simple


Click Define

In Define Simple Boxplot: Summaries for Groups of Cases window:

Select bmi and move it to Variable box

Select smoking and move it to Category Axis box

Click OK



8.7 Error Bar (please use data entry spss.sav file!)

Suppose we wish to compare between mean of BMI (kg/m2) of smoking and non-smoking

mothers

Click Graphs


Select Error Bar…


In Error Bar window:

Select Simple


Click Define

In Define Simple Error Bar: Summaries for Groups of Cases window:

Select bmi and move it to Variable box

Select smoking and move it to Category Axis box

Click OK



8.8 Scatter/Dot

Suppose we wish to know the correlation between weight (kg) and height (cm) of the

mothers

Click Graphs


Select Scatter/Dot…


In Scatter/Dot window:

Select Simple Scatter

Click Define

In Simple Scatterplot window:

Select weight variable and move it into Y axis box

Select height variable and move it into X axis box

Click OK




9. Opening an existing data in SPSS file or SPSS output

9.1 When starting SPSS program

9.1.1 SPSS file

Select Open an existing data source

Select the intended file, let‟s say measles.sav or More Files… to select another file

Click OK


9.1.2 SPSS output

Select Open another type of file

Select the intended file or More Files… to select another file

Click OK

9.2 From the Window Explorer

9.2.1 SPSS file


Double click on the intended file

9.2.2 SPSS output


Double click on the intended file

INTRODUCTION OF SPSS 15.0 FOR WINDOWSLab Activity Community Research Program IV STUDENT’S GUIDE INTRODUCTION OF SPSS 15.0 FOR WINDOWS Faculty of Medicine Universitas Padjadjaran

Documents