Step 1:
Analyze > Compare Means > Paired Samples T Test

IV = Training (levels: before training; after training)

DV = Test Score
Step 2
In the PairedSamples T Test dialogue box, move the two variables for comparison into the "Paired Variables" box.
You shouldn't need to make any changes unless you want to change the confidence level limits or exclude cases (via the "Options" button).
Click "OK" to get the output.
Step 3: Look at the Output
The two main tables you need to look at are the "Paired Samples Statistics" table and "Paired Samples Test" table.
The means (and standard deviations) in the "Paired Samples Statistics" table can be used in the writeup of the descriptive statistics and calculating the effect size.
The last three columns of the "Paired Samples Test" table display the results of the ttest, where "t' is the tvalue, "df" is the Degrees of Freedom, and "Sig. (2tailed)" is the significance level.
Nonparametric Ttests
A ttest compares the means of two groups.
There are two types:

Paired/Dependent  for withinsubjects

Unpaired/Independent  for betweensubjects
Assumptions:

The dependent variable (DV) should be scale data (either interval or ratio)

The independent variable (IV) should consist of two categorical groups

No obvious outliers

The data is roughly normally distributed for each group of the independent variable
1. Paired ttest (example)
Step 4: Calculate the Effect Size
To calculate the effect size (Cohen's d), you need the means (x) and standard deviations (SD).
d = (x1  x2) ÷ ((SD1 + SD2) ÷ 2)
Very Small < 0.3
SmallMedium > 0.3; < 0.5
MediumLarge > 0.5; < 0.8
Large > 0.8
Step 5: Writeup
[Error Graph]
Figure 1. Mean Test Score and Confidence Intervals for Before and After Training
On average, participants had lower test scores before training than after training. The mean test score before training was 43.4 compared to 54.5 after training. The confidence intervals show that the means are reasonably close to the population mean. *Report upper and lower bounds (if required) from Analyze > Explore ...*
A paired samples ttest (twotailed) revealed a significant difference in test scores at the 5% level, t (25) = 57.832, p < 0.001, with a significantly higher test score after training than before training. The mean difference between the conditions was 11.08, with a mediumlarge effect size (d = 0.55).
*You might also want to report from the "Paired Samples Test" table*
The 95% confidence intervals for the estimated population mean difference is between [lower] and [upper].
2. Independent ttest (example)
Step 1:
Analyze > Compare Means > IndependentSamples T Test

IV = Gender (levels: male, female)

Male labelled "1"

Female labelled "2"


DV = Height in centimeters
Step 2
In the IndependentSample T Test dialogue box, move the IV to the "Grouping Variable" box and the DV to the "Test Variable(s)" box.
You will need to define the levels of the grouping variable before contnuing. To do this, you need to select the grouping variable and then click "Define Groups" and define "Group 1" as "1" (the label for "Male") and "Group 2" and "2" (the labale for female).
N.B. If there had been a third group (or fourth, fifth, etc.), e.g., "unknown" labelled "3", you could compare groups "1" and "3" instead (or "2" and "3"). This is why you're asked to define the groups  incase you have more than 2. You can still only test two groups, but you need to say which two.
You shouldn't need to make any other changes unless you want to change the confidence level limits or exclude cases (via the "Options" button).
Click "OK" to get the output.
Step 3: Look at the Output
SPSS provide you with the two tables you need for reporting the outcome.
The means (and standard deviations) in the "Group Statistics" table can be used in the writeup of the descriptive statistics and for calculating the effect size.
The third, fourth and fifth columns of the "Independent Samples Test" table display the results of the ttest, where "t' is the tvalue, "df" is the Degrees of Freedom, and "Sig. (2tailed)" is the significance level.
Step 4: Calculate the Effect Size
To calculate the effect size (Cohen's d), you need the means (x) and standard deviations (SD).
d = (x1  x2) ÷ ((SD1 + SD2) ÷ 2)
Very Small < 0.3
SmallMedium > 0.3; < 0.5
MediumLarge > 0.5; < 0.8
Large > 0.8
Step 5: Writeup
[Error Graph]
Figure 1. Mean Height (cm) and Confidence Intervals for Men and Women
On average, women were shorter than men. The mean height for women was 165.75cm compared to 175.125cm for men. The confidence intervals show that the means are reasonably close to the population mean. *Report upper and lower bounds (if required) from Analyze > Explore ...*
An independent samples ttest (twotailed) revealed a significant difference in test scores at the 5% level, t (30) = 2.395, p = 0.023, with a significantly higher height for men than for women. The mean difference between the conditions was 9.375, with a large effect size (d = 0.85).
*You might also want to report from the "Independent Samples Test" table*
The 95% confidence intervals for the estimated population mean difference is between [lower] and [upper].