You have a research study where you measure the same group of people two different times. Maybe you gave them a test, gave them a training course, and then gave them the same test again. You want to know if their scores changed. The best way to find out is by using a paired-samples t-test.
At SPSS Solutions, we often see students get confused about when to use this test. It is only for when you have two scores from the same person or matched pairs. If you have two completely different groups of people, you should use an independent-samples t-test instead.
When to Use a Paired-Samples T-Test
You use this test when your data fits a specific design. First, you must have two continuous variables. These are variables that use numbers, like test scores or blood pressure readings. Second, each case in your dataset must have a pair of scores. This usually happens in pre-test and post-test studies.
You should also think about your assumptions. The difference between the two scores should be roughly normally distributed. If your data has massive outliers or is very skewed, the results might not be reliable. You can check this by looking at a histogram of the difference between your two variables.
How to Perform the Test in SPSS
Open your dataset in SPSS. Make sure your two variables are already in the system. Go to the top menu and click on Analyze. Then, move your mouse to Compare Means and select Paired-Samples T Test.
A box will pop up. You need to select your two variables. Click on the first variable, then hold the Ctrl key on your keyboard and click the second variable. Click the arrow button to move them into the Paired Variables box. Once they are there, click OK.
If you have multiple pairs, you can move all of them at once. SPSS will run a separate t-test for every pair you include in the box.
Understanding the SPSS Output
SPSS will show you three main tables. The first table is the Paired Samples Statistics. This gives you the mean, the number of people, and the standard deviation for both variables. Check this to make sure your data looks correct.
The second table is the Paired Samples Correlations. This shows you if the scores from the first time are related to the scores from the second time. If the correlation is high, it means people who scored high first often scored high again.
The third table, Paired Samples Test, is the one you really need. Look at the column labeled Sig. (2-tailed). This is your p-value. If this number is less than .05, it means there is a statistically significant difference between your two points in time. If it is higher than .05, you do not have enough evidence to say the scores changed.
How to Report Your Results in APA Style
When you write your results, you need to include the t-statistic, the degrees of freedom, and the p-value. APA 7th edition has specific rules for this.
You should report it like this: “A paired-samples t-test showed a significant difference between pre-test scores (M = 75.20, SD = 5.10) and post-test scores (M = 82.50, SD = 4.80), t(29) = 4.35, p < .001."
Notice how we use italics for the letters t, M, SD, and p. Also, notice that we did not write p = .000. In statistics, we use p < .001 when the output shows .000 because a p-value can never actually be zero.
Common Mistakes to Avoid
One common mistake is forgetting that this test is for related samples. If you accidentally use data from two different groups, your results will be wrong. Always double-check your data layout in Variable View to ensure your variables are numeric.
Another mistake is focusing only on the p-value. A result can be statistically significant even if the change is very small. Always look at the means in the first table to see if the change is actually meaningful for your research. Sometimes, you might find a significant change in scores, but the difference is only one or two points, which might not matter in the real world.
If your dataset is missing a score for one of the two time points, SPSS will exclude that person from the analysis for that pair. This is called listwise deletion. You should check your descriptive statistics to see if you lost many participants due to missing data.
Helpful Resources
If you are working with a more complex design, like having more than two time points, a t-test will not work. You might need to look into SPSS ANOVA without raw data if you are trying to compare group averages using only summary statistics, though for repeated measures, you would generally need the full dataset for a repeated-measures ANOVA.
Always keep your syntax if you plan to run these tests again. Saving your syntax allows you to quickly repeat your analysis if you decide to change your data or add more participants later. It acts as a permanent record of what you did in your study.
FAQs
- What does the t-value mean? The t-value shows how far apart the two means are compared to the spread of the data. A larger t-value usually means a stronger difference.
- What if my p-value is .06? In many fields, .06 is considered not significant because it is above the standard .05 limit. You should be careful about claiming a change happened if you do not meet the .05 threshold.
- Do I need to report the standard error? You do not have to, but it is often helpful. Most researchers prefer to report the mean and standard deviation.
If you ever feel stuck, remember that running the analysis is only half the work. The other half is making sure your data is clean before you start. Check for typos and errors in your data first. If your data is clean, the test results will be much easier to trust.