SPSS Guide
Topic Performing Tests in SPSS
Notes
Taking a Random Sample From the Population
Random Sample From Population
Step 1: Sort Cases
Data —> Sort Cases
place the variable “….” into the “Sort By” window, select “Sort Order:
Descending”, and click OK.
Step 2: Select Cases
Data —> Select Cases
Select “Random Sample of Cases” and click on the “Sample” button.
Another window should pop up – here, select and type “Exactly … cases
from the first …. cases”.
Confidence Intervals
Calculating C.I.
Analyze → Compare Means → Mean
Place the variable “Southgate Group Reading Test Score” into the
“Dependent List”
Click “Options” and select the cell statistics Mean, Number of Cases,
Standard Deviation, and Std Error of Mean.
Use the output to estimate the confidence interval
Graphical Display
Chart Builder —> Bar Chart
SPSS Guide 1
, in “Element Properties” on the right, select Statistic: Mean, tick Display Error
Bars, and select Confidence Intervals, Level: 95
Exploring Means
Comparing subgroup means
Analyze → Compare Means → Means
Place one variable (the outcome variable) in the Dependent List and the
independent variable (e.g. sex of the child) in the Independent List (“Layer 1
of 1”)
Exploring Data
Exploring Descriptives and plots
Analyze → Descriptive Statistics → Explore
Put the variable “….” in the Dependent List
under the option “Statistics”, select Descriptives (with Confidence Interval for
Mean at 95%), and Outliers;
under the option “Plots”, select Boxplot (with Factor levels together), Stem-
and-Leaf, and Histogram.
Plots —> Dependents together — Choose dependents together in the plot
section if you want two variables to be displayed in the same graph
SPSS Guide 2
, Exploring two variables with different missing values
Analyze → Descriptive Statistics → Explore
Put the variables in the Dependent List
Options → Exclude cases pairwise
If the default “Exclude cases listwise” is selected, then the number of valid
cases will be the same for both variables - cases with missing data on either
variable will not be considered
Defining Outlier value as missing
syntax: missing values [variable name] (value)
variable view —> missing column —> click the three dots —> Define the missing
value
T-test
Single Sample t-test
We have to know the normative population mean (e.g. μ=100 for IQ tests)
Analyze → Compare Means → One-Sample T-test
Place “variable” into the “Test Variable(s)” box — define the Test Value.
SPSS Guide 3
Topic Performing Tests in SPSS
Notes
Taking a Random Sample From the Population
Random Sample From Population
Step 1: Sort Cases
Data —> Sort Cases
place the variable “….” into the “Sort By” window, select “Sort Order:
Descending”, and click OK.
Step 2: Select Cases
Data —> Select Cases
Select “Random Sample of Cases” and click on the “Sample” button.
Another window should pop up – here, select and type “Exactly … cases
from the first …. cases”.
Confidence Intervals
Calculating C.I.
Analyze → Compare Means → Mean
Place the variable “Southgate Group Reading Test Score” into the
“Dependent List”
Click “Options” and select the cell statistics Mean, Number of Cases,
Standard Deviation, and Std Error of Mean.
Use the output to estimate the confidence interval
Graphical Display
Chart Builder —> Bar Chart
SPSS Guide 1
, in “Element Properties” on the right, select Statistic: Mean, tick Display Error
Bars, and select Confidence Intervals, Level: 95
Exploring Means
Comparing subgroup means
Analyze → Compare Means → Means
Place one variable (the outcome variable) in the Dependent List and the
independent variable (e.g. sex of the child) in the Independent List (“Layer 1
of 1”)
Exploring Data
Exploring Descriptives and plots
Analyze → Descriptive Statistics → Explore
Put the variable “….” in the Dependent List
under the option “Statistics”, select Descriptives (with Confidence Interval for
Mean at 95%), and Outliers;
under the option “Plots”, select Boxplot (with Factor levels together), Stem-
and-Leaf, and Histogram.
Plots —> Dependents together — Choose dependents together in the plot
section if you want two variables to be displayed in the same graph
SPSS Guide 2
, Exploring two variables with different missing values
Analyze → Descriptive Statistics → Explore
Put the variables in the Dependent List
Options → Exclude cases pairwise
If the default “Exclude cases listwise” is selected, then the number of valid
cases will be the same for both variables - cases with missing data on either
variable will not be considered
Defining Outlier value as missing
syntax: missing values [variable name] (value)
variable view —> missing column —> click the three dots —> Define the missing
value
T-test
Single Sample t-test
We have to know the normative population mean (e.g. μ=100 for IQ tests)
Analyze → Compare Means → One-Sample T-test
Place “variable” into the “Test Variable(s)” box — define the Test Value.
SPSS Guide 3
