INTRODUCTION TO SPSS
SPSS for Windows Commands
To CREATE an SPSS data file (primary data--not "borrowed"):
- GO TO VARIABLE VIEW TAB AT BOTTOM OF SCREEN
- Give the variable a NAME that has 8 letters or less
- Give the variable a LABEL (a statement in English that
describes the variable)
- Give the variable VALUES
- Give values (codes) and value labels to each of your codes
- ADD each of the values and value labels
- CONTINUE
- OK
- Give the variable MISSING VALUE codes
- GO TO THE DATA VIEW TAB AT THE BOTTOM OF
SCREEN
- Highlight the first column/first row; type the data; click
- Highlight the second column/first row; type the data; click
- Etc.
To SAVE the SPSS data file (primary data) that you have created:
- FILE
- SAVE AS
- Give the datafile a name (the file extension is ".sav")
To PRINT data dictionary information from your file.
- FILE
- DISPLAY DATA FILE INFORMATION (from working file)
- FILE
To OPEN an already existing ("borrowed") SPSS data file:
- FILE
- OPEN
- Drive: A OR
- C OR
- file server (CHSBS1)
- File type: SPSS for DOS--SPSS/PC+ (*.sys) OR SPSS for Windows (*.sav)
- File name: "Name of file"
To OPEN an already existing (you created it earlier) SPSS output file saved on your
disk or hard drive:
- FILE
- OPEN
- Drive: A OR C
- File type: SPSS document that ends .spo
- File name: "Name of file"
To CHECK the meaning of the codes of an existing variable:
- OPEN your SPSS datafile (see above)
- GO TO VARIABLE VIEW TAB (bottom of screen)
- GO TO VALUES
- Click on the three dots (...) and the value
- label box opens.
ALTERNATIVE way to CHECK the meaning of the codes of an existing variable
OPEN your SPSS datafile (see above)
UTILITIES
VARIABLES
Click on the variable in the dataset that interests you (SCROLL down
until you find it)
To create FREQUENCY and PERCENTAGE distributions:
- OPEN your SPSS data file (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- FREQUENCIES
- Highlight the first variable that interests you; click on the arrow
(>)
- Highlight the second variable that interests you; click on the arrow
(>)
To create MEASURES OF CENTRAL TENDENCY and MEASURES OF DISPERSION
- OPEN your SPSS data file (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- FREQUENCIES
- Highlight the first variable that interests you; click on the arrow
(>)
- Highlight the second variable that interests you; click on the arrow
(>)
- Click on STATISTICS
- Click the boxes for the statistics you want
To create just the MEAN and MEASURES OF DISPERSION
- OPEN your SPSS datafile (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- DESCRIPTIVES
- Highlight the first variable that interests you; click on the arrow
(>)
- Highlight the second variable that interests you; click on the arrow
(>)
To create Z-scores (standardized scores)
- OPEN your SPSS data file (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- DESCRIPTIVES
- Highlight the variable that interests you; click on the arrow
(>)
- Click on the little box on the left "Save
standardized score"
To create MEASURES OF CENTRAL TENDENCY and MEASURES OF DISPERSION for ONE
VARIABLE for Subgroups (for example, Number of Hours Worked for Men and
for Women)
- OPEN your SPSS data file (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- EXPLORE
- Highlight the variable that interests you; move it to the
Dependent List Box
- Highlight the variable that divides your dataset into subgroups;
move it to the Factor List box
- Click on STATISTICS
- Click the boxes for the statistics you want
To create UNIVARIATE GRAPHS
- OPEN your SPSS data file (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- FREQUENCIES
- Highlight the first variable that interests you; click on the arrow
(>)
- Highlight the second variable that interests you; click on the arrow
(>)
- Click on CHARTS
- Click the boxes for the graphs you want
MORE on CREATING UNIVARIATE GRAPHS
OPEN your SPSS datafile (see above)
GRAPH
, then LEGACY DIALOG
BAR
SIMPLE
DEFINE
Highlight the variable that interests you; move it to the Category Axis
box
UNDER OPTIONS: Eliminate Missing Data from your graph
OK
To create BOX PLOTS
- OPEN your SPSS data file (see above)
- GRAPHS
, then LEGACY DIALOG
- BOX PLOT
- Select the option SIMPLE, then click on DEFINE
- Highlight the variable that interests you and move it to the VARIABLE
BOX
- Highlight the variable that divides your dataset into subgroups and
move it to the CATEGORY AXIS BOX
To create CROSSTABULATIONS
- OPEN your SPSS datafile (see above)
- ANALYZE
- DESCRIPTIVE STATISTICS
- CROSSTABS
- Highlight your dependent variable; click on row arrow (>)
- Highlight your independent variable; click on column arrow (>)
- CELLS
- Click on COUNTS--OBSERVED
- Click on PERCENTS--COLUMN
- STATISTICS
To create CORRELATION COEFFICIENTS
- OPEN your SPSS datafile (see above)
- ANALYZE
- CORRELATE
- BIVARIATE
- Highlight the first variable that interests you; click on the arrow
(>)
- Highlight the second variable that interests you; click on the arrow
(>)
- Under CORRELATION COEFFICIENTS, click on PEARSON
To SAVE an SPSS output file to your disk:
- In the OUTPUT window--
- SAVE AS
- Drive: A
- Give the datafile a name (It will end with the file extension .spo)
To MOVE tables from SPSS into WORD PERFECT or WORD:
- In the OUTPUT window--
- Click somewhere inside the table you want to move
- EDIT
- COPY (or CUT)
- MINIMIZE the OUTPUT window and MINIMIZE anything else that is open.
(Minimize is the left-hand button at the top right side)
- Now, get into WORD PERFECT or WORD the way you usually do--through
- START or the ICON on the desktop--
- EDIT
- PASTE SPECIAL
- SPSS Pivot Table or PICTURE (depending on word processor)
- Click somewhere inside the table and then move it to where you want it
on the page
To INSERT words (like Table titles and Source information) into the SPSS output window:
To PRINT anything:
To RECODE variables
- OPEN your SPSS datafile (see above)
- TRANSFORM
- RECODE
- INTO SAME VARIABLE or INTO DIFFERENT VARIABLE
- Click on first variable you want to recode
- NOTE: If you choose "recode to different variable," then you
must give the new variable a name and then click CHANGE
- OLD AND NEW VALUES
- Type old value, then move over and type new value
- Type old value, then move over and type new value
- Type old value, then move over and type new value
To CREATE INDICES or SCALES
OPEN your SPSS datafile (see above)
TRANSFORM
COMPUTE
Give your index or scale a name (Target Variable Name)
Highlight the first variable to be part of your index/scale; move it to
the right box. Add a plus (+) sign.
Highlight the second variable to be part of your index/scale; move it
to the right box. Add a plus (+) sign.
Highlight the third variable to be part of your index/scale; move it to
the right box. Add a plus (+) sign.
Continue until your index/scale is complete.
OK
To COMPUTE CRONBACH's ALPHA
OPEN your SPSS datafile (see above)
ANALYZE
SCALE
RELIABILITY
Move each variable that is part of your scale (or your proposed scale)
from the left-hand box to the right-hand box.
OK
To COMPARE PRE-TEST MEANS and POST-TEST MEANS
OPEN your SPSS datafile (see above)
ANALYZE
COMPARE MEANS
PAIRED-SAMPLE T TEST
Click on two variables (your pre-test and your post-test variables)
OK
To COMPARE EXPERIMENTAL GROUP MEANS and CONTROL GROUP MEANS
OPEN your SPSS datafile (see above)
ANALYZE
COMPARE MEANS
INDEPENDENT SAMPLE T TEST
Click on variable of interest (either your pre-test or your post-test
variable). Move to right.
Click on variable that distinguishes between your experimental and
control group. Move into area labeled GROUP.
Click DEFINE GROUPS. Indicate the CODE for the experimental
group in Group 1 and the CODE for the control group in Group
2.
OK
CREATING BIVARIATE GRAPHS--Bar Graph
OPEN your SPSS datafile (see above)
GRAPH
, then LEGACY DIALOG
BAR
CLUSTERED
DEFINE
Highlight the dependent variable that interests you; move it to the
Category Axis box
Highlight the independent variable that interests you; move it to the
"Define Clusters by" box
UNDER OPTIONS: Eliminate Missing Data from your graph
OK
CREATING BIVARIATE GRAPHS--SCATTERGRAM (Scatterplot)
OPEN your SPSS datafile (see above)
GRAPH, then LEGACY DIALOG
SCATTER
SIMPLE
DEFINE
Highlight the dependent variable; move it to the Y-axis box
Highlight the independent variable; move it to the X-axis box
OK
To add the regression line: Double click on the graph (to get to editor
mode), then
CHART
OPTIONS
Check FIT LINE
OK
Return to the Main Page
for SOC 200