SPSS for dummies

Lesson 1 (DATA ENTRY, SELECT CASES, FREQUENCY TABLE)

1.2 STARTING SPSS - I assume yall know how
1.3 CREATING DATA FILE - I assume yall know how :)
1.4 SPSS Data formats - got 2 types
a. First format is used when the data consist of several scores for each individual.
b. The second format is used for data from an independent-measures study using a
separate group of participants for each treatment condition. This kind of data is
entered into the data matrix in a stacked format.
1.5 DEFINE THE VARIABLES
a. Naming variable
- Click on variable view then double click on the first cell below the Name
column then you type the variable in then you define the variable
properties
b. How to define variable properties
- You need to define it before you enter data records
● Type: String (for text), Numeric (for numbers). The default setting is Numeric.
● Width: Size of variable. The default setting is 8
● Decimals: No of decimal places. The default setting is 2
● Label: Description of the variable
● Values: Code values for the variables (e.g. 1=female, 2=male)
● Measure: Scale (for interval data – this is the default value for numeric
variables) Ordinal (for ordinal data)
Nominal (for nominal data)
1.6 ENTER DATA
- Just go to data view tab and enter the data lol
1.7 SAVE THE FILE - assuming yall know how
1.8 SELECT CASES
- SPSS Functions Menu , Select Data → Select Cases → click satisfied → If →
Enter the condition (s) → Continue → OK
1.9 HOW TO CONSTRUCT A FREQUENCY TABLE
- SPSS Functions Menu , Select Analyze → Descriptive Statistics Frequencies →
enter the variable(s) → OK
1.10 HOW TO CONSTRUCT A TWO-WAY TABLE (CROSS TABULATION)
- From SPSS Functions Menu , Select Analyze → Descriptive Statistics →
Crosstabs → Row(s ): click the variable(s)
Column(s ): click the variable(s) → OK
Lesson 2 DESCRIPTIVES, COMPUTE, HISTOGRAM
2.3 HOW TO USE DESCRIPTIVE STATS TO DESCRIBE DATA
SPSS STEPS
- Select Analyze → Descriptive Statistics → Descriptives → click on Options
….→select mean, standard deviation, maximum, minimum, skewness →
Continue → OK
Example: Find the max, min, std dev, mean, and skewness of sleep hours

Steps:

Results:
2.4 HOW TO CREATE A NEW VARIABLE (COLUMN) USING ‘COMPUTE’
SPSS STEPS
- Select Transform → Compute Variable….→Enter the new variable name
→Enter the formula → OK
Example: difference between sleep last night and sleep last 7 nights
Steps:

* make sure to type in the “target variable” to proceed

Results: your new variable is in the variable tab (scroll down, it should be the latest one
added in)

* you can compute a descriptive statistic or any tables or graphs using the new variable
2.5 HOW TO PRESENT DATA USING HISTOGRAM
SPSS steps:
- Select Graphs → Histogram → Enter the variable name → OK
Example : histogram for studying hours

- In the results, you can double click on the chart to change the title and label the
axis accordingly
Lesson 3 RECODE, PIE CHART, BAR CHART
3.2 HOW TO CREATE A NEW VARIABLE (COLUMN) USING ‘RECODE’
- Useful for grouping records

Then you’ll see;

3.3 GRAPHS
a. Pie chart
SPSS steps:
- Select Graphs → Legacy Dialogs → Pie → Summarize for Groups of Cases
→Click Define → Click N of cases→Define Slices by: (choose the variable)
- In the results:
→ Click Titles… button→ Enter Title, Subtitle and Footnote → Continue → OK
*If you want to customize the labels for each slice to include the values or
percentages: Double click on the pie chart that is displayed to get the Chart
Editor menu then choose Elements→ Show Data Labels. On the Properties
window, click on Percentages → OK
b. Simple Bar Chart
SPSS Steps:
- Select Graphs →Legacy Dialogs→ Bar → Simple, click on Define → Choose N
of cases → enter the variable name in Category Axis → Click Titles… button→
Enter Title, Subtitle and Footnote → Continue → OK
- To modify the chart:
Double-Click on the bar chart, to get the Chart Editor Select the options or the
Elements from the Chart Editor menu
*You can interactively make changes to the chart by double-clicking on the
appropriate item on the chart. E.g. click on the x-axis to change the x-axis label.
c. Clustered bar chart
SPSS steps:
- Select Graphs →Legacy Dialogs→ Bar → Clustered, Click Define → Enter the
variable name in Category Axis and Enter another variable name in Define
Clusters By → OK
d. Stacked Bar Chart
SPSS steps:
- Select Graphs → Legacy Dialogs → Bar → Stacked → Click Define → Enter the
variable name in Category Axis and Enter another variable name in Define
Stacks By → OK
3.4 HOW TO SORT CASES
SPSS Steps:
- Data → Sort Cases → Enter and the variable names in Sort By → Select Sort
order → OK
Lesson 4: Confidence Interval
- Analyze → Descriptive Statistics → Explore
- Dependent list = put the variable (eg. salary)
- Factor list (if applicable) = put the categories (eg. male, female)
- Statistics → indicate % confidence level

Lesson 5: One Sample t-test

- SPSS steps:
1. Analyze → Compare Means → One-Sample t-Test
2. Test variables → select the variable to be tested (eg. hours of sleep)
3. Test Value → enter the value that you specified in your H0 (eg. 10 hours)
4. Check % confidence level

- Written steps:
1. List out hypothesis and specify the claim
H0 : µ = 0
H1 : µ </> 0 (claim?)
2. List t-value
3. List p-value (aka Sig.)
4. Reject H0 if p-value < α

Note:
● If it is a two-tail test, take p-value directly from Sig.
● If it is a RIGHT-tail test (H1: μ > 10)
○ If t-value is positive, p-value = sig. / 2
○ If t-value is negative, p-value = 1 - (sig. / 2)

● If it is a LEFT-tail test (H1: μ < 10)

○ If t-value is positive, p-value = 1 - (sig. / 2)
○ If t-value is negative, p-value = sig. / 2
Lesson 6
Independent Samples t-test
- For 2 populations (eg. men, women)
- Test variable (eg. salary)

- SPSS steps:
1. Analyze → Compare Means → Independent Samples t-Test
2. Choose Test Variable
3. Define groups (choose the 2 populations)
4. Check % confidence level

- Written steps:
(Part A)
1. List out hypothesis for variance homogeneity
H0 : σ2 A = σ2 B
H1 : σ2 A ≠ σ2 B

2. F value = (check SPSS table)

3. p-value for F-test = (check SPSS table)
4. α = (check significance level)
5. Reject H0 if p-value < α ; variances are not equal
(Part B)
1. List out hypothesis to test means difference & specify the claim
H0 : µ1 − µ2 = 0 (claim)
H1 : µ1 − µ2 ≠ 0

(or use “µ1 − µ2 < / > 0” 0 if it is one-tail)

2. t-value = (check SPSS table)

3. p-value for t-test = (check SPSS table, REMEMBER to divide by two if the
question is a one-tail test)
4. Reject H0 if p-value < α ; claim is true / false ; …
Paired-Samples t-test
- for “before & after” scenarios

- SPSS steps
1. Analyze → Compare Means → Paired-Samples t-Test
2. Set Variable 1 = Before ; Variable 2 = After
3. Check % confidence level

- Written steps:
1. List out hypothesis to test means difference & specify the claim
H0 : µD = 0 (claim?)
H1 : µD ≠ 0
2. List out p-value (aka Sig.)
3. List out α
4. Reject H0 if p-value < α

Lesson 7 (ANOVA)

7.3 HOW TO DO ANOVA TEST

From SPSS Functions Menu,
Select Analyze → Compare Means → One-Way ANOVA → Dependent List:
...Choose the variable (e.g. age)
Factor: ....Select the grouping variable (e.g. by gender)
→ Click Options
In the One-Way ANOVA Options window, Select Descriptive and Means Plot
→ Click Continue→ OK

SPSS will display 3 outputs:

• Descriptive statistics of the variable for each group
• ANOVA output (see sample below)
• Line plot showing the mean values for the variable for each group
SPSS Version 20 LAB MANUAL FOR PSY223S Page 2 of 3
• OUTPUT
7.4 HOW TO REPORT THE RESULTS
AN EXAMPLE: A one-way between groups ANOVA was performed to compare the
impact of age of job satisfaction. Participants were divided into three groups based on
their age groups: (Group 1:29 years or less. Group 2: 30 – 44 years, and Group 3: 45
years and above). The outcome variable was found to be normally distributed and equal
variance are assumed based upon the results of Levene’s test F(432)=.746, p=.475).
There was a statistically significant difference in job satisfaction scores for the three age
grops (F2, 432)=4.6, p=.01). The magnitude of the difference in means and the effect
size was very small (partial eta squared = 0.02). Post hoc scores for Group 1 (M=21.36,
SD=4.55, 95% CI=20.6, 22.1) was significantly different from Group 3 (M=22.96,
SD=4.49, 95% CI=22.2, 23.7. Group 2 (M=22.1, SD=4.15, 95% CI=21.4,22.7) did not
differ significantly from either Group 1 or 3.

Lesson 9 (Two factor ANOVA)

Analysis of Variance (ANOVA) test is used to test the differences in the means for two
factors and
the interaction. Three hypotheses are tested i.e.
I. For factor A:
𝐻0: 𝜇1 = 𝜇2 = 𝜇3 = ⋯ . = 𝜇𝑛
𝐻1: 𝑎𝑡 𝑙𝑒𝑎𝑠𝑡 𝑜𝑛𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑒𝑎𝑛𝑠 𝑎𝑟𝑒 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑡

II. For factor B:

𝐻0: 𝜇1 = 𝜇2 = 𝜇3 = ⋯ . = 𝜇𝑛
𝐻1: 𝑎𝑡 𝑙𝑒𝑎𝑠𝑡 𝑜𝑛𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑒𝑎𝑛𝑠 𝑎𝑟𝑒 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑡

III. For interaction:

𝐻0: 𝑇ℎ𝑒𝑟𝑒 𝑖𝑠 𝑛𝑜 𝑖𝑛𝑡𝑒𝑟𝑎𝑐𝑡𝑖𝑜𝑛 𝑒𝑓𝑓𝑒𝑐𝑡 𝑓𝑜𝑟 𝑓𝑎𝑐𝑡𝑜𝑟𝑠 𝐴 𝑎𝑛𝑑 𝐵

𝐻1: 𝑇ℎ𝑒𝑟𝑒 𝑖𝑠 𝑎𝑛 𝑖𝑛𝑡𝑒𝑟𝑎𝑐𝑡𝑖𝑜𝑛 𝑒𝑓𝑓𝑒𝑐𝑡 𝑓𝑜𝑟 𝑓𝑎𝑐𝑡𝑜𝑟𝑠 𝐴 𝑎𝑛𝑑 𝐵

9.3 HOW TO ENTER DATA FOR TWO-FACTOR ANOVA TEST

1. The scores are entered into the SPSS data editor in a stacked format, which means
that all the scores from all the different treatment conditions are
entered in a single column (VAR00001).

2. In a second column (VAR00002) enter a code number to identify the level of factor A
for each score. If factor A defines the rows of the data matrix, enter
a 1 beside each score from the first row, enter a 2 beside each score from the
second row, and so on.

3. In a third column (VAR00003) enter a code number to identify the level of

factor B for each score. If factor B defines the columns of the data matrix,
enter a 1 beside each score from the first column, enter a 2 beside each score
from the second column, and so on. Thus, each row of the SPSS data editor will have
one score and two code numbers, with the score in the first column, the code for factor
A in the second column, and the code for factor B in the third column.

9.4 HOW TO DO ANOVA TWO-FACTOR TEST

From SPSS Functions Menu,
Select Analyze → General Linear Model → Univariate
In the Univariate window.
→Highlight the column label for the set of scores (weight loss) in the left
box and click the arrow to move it into the Dependent Variable box.
→ One by one, highlight the column labels for the two factor codes and
click the arrow to move them into the Fixed Factors box. (e.g. diet and
exercise level)

→ If you want descriptive statistics for

each treatment, click on
the Options box, select Descriptives,
and click Continue.
→ If you want plot the marginal means,
click on the Plot box, select a
variable (factor A) for the Horizontal
Axis and another variable
(factor B) for the Separate Lines , and
click Add.
→Click OK

SPSS Output
Part of the program output is shown below. The output begins with a table listing the
factors (not shown below), followed by a table showing descriptive statistics including
the mean and standard deviation for each cell or treatment condition. The results of the
ANOVA are shown in the table labeled Tests of Between-Subjects Effects. The top row
(Corrected Model) presents the between-treatments SS and df values. The second row
(Intercept) is not relevant for our purposes. The next three rows present the two main
effects and the interaction (the SS, df, and MS values, as well as the F-ratio and the
level of significance), with each factor identified by its column number from
the SPSS data editor. The next row (Error) describes the error term (denominator of the
F-ratio), and the final row (Corrected Total) describes the total variability for the entire
set of scores. (Ignore the row labeled Total.)
HOW TO REPORT THE RESULTS

Lesson 10 (CORRELATION AND LINEAR REGRESSION ANALYSIS)

10.2 INTRODUCTION
In this lesson, we can consider the relationship between two variables in two ways:

(a) Correlation Analysis

To measure the strength of relationship between two variables (i.e. by finding the
coefficient of correlation and the coefficient of determination)

(b) Regression Analysis

To estimate the value of a dependent variable (Y) on the basis of an independent
variable (X), by finding the equation of the linear regression line.

10.3 HOW TO OBTAIN A SCATTER DIAGRAM

• A plot of paired observations is called a scatter diagram.
• To do a Scatter Plot of Y on X:

From SPSS Functions Menu,

Select Graphs → Scatter..... → Simple → Define:
Y axis : ...enter the dependent variable
X Axis: ....enter the independent variable
→ OK

10.4 HOW TO OBTAIN THE CORRELATION COEFFICIENT

From SPSS Functions Menu,
Select Analyze → Correlate → Bivariate → Select the two variables that you wish to
correlate →OK
The following Correlation output will be displayed:

(a) Comment/ Interpretation of the value of r

10.5 HOW TO DRAW A REGRESSION LINE

• To draw a regression line:
Draw a scatter plot then Double click on the scatter diagram.
Select Chart → Options→ The Scatterplot Options window is displayed:
Click Total in the Fit Line
→ OK
SPSS Version 20 LAB MANUAL FOR PSY223S Page 3 of 5
10.6 HOW TO OBTAIN THE LINEAR REGRESSION EQUATION
(a) Do a linear Regression: Y on X
From SPSS Functions Menu,
Select Analyze→ Regression → Linear
Dependent: the dependent variable (y)
Independent: the independent variable (x)
→ OK
(b) To obtain the linear regression coefficients and equation, refer to the linear
regression
Output:

10.7 HOW TO OBTAIN THE COEFFICIENT OF DETERMINATION

(a) To obtain the coefficient of determination
Refer to linear regression output:

The coefficient of determination is the value of R Square ( = r2)

(b) Comment / Interpretation of the values of coefficient of determination

r2 x 100% of the variation of y is explained by the variation of x: