How to Run and Interpret a Pearson Correlation in SPSS

When you want to see if two variables move together, you need a correlation. A Pearson correlation checks the strength and direction of a linear relationship between two continuous variables. If one variable goes up, does the other go up too? Or does it go down?

At SPSS Solutions, we get questions every day about how to run this test and what the output actually means. Let us walk through the whole process step by step so you can finish your homework or research paper with confidence.

When to Use a Pearson Correlation

You use a Pearson correlation when your data meets specific rules. If you break these rules, your numbers will not mean anything. Here is what you need:

  • You have two variables.
  • Both variables use a continuous scale. This means interval or ratio data, like test scores, height, or weight.
  • Your data has a linear relationship. You can check this easily by using an SPSS Scatterplots & Fit Lines Tool before you run the main number crunching.
  • Your data follows a normal distribution.

Do not use this test if your data is categorical or ranked ordinal data. For ranks, you should use Spearman’s rank correlation instead.

Assumptions to Check First

Never run a statistical test without checking your assumptions. For a Pearson correlation, you need to check three main things:

  • Measurement level: Make sure both variables are set to Scale in the Variable View tab.
  • Normality: Your variables should form a normal distribution curve. You can test this in SPSS by running a Shapiro-Wilk test through the descriptive statistics menu.
  • Linearity: The relationship between your variables must be a straight line, not a curved one. A simple scatterplot will show you if this is true.

Step-by-Step Instructions in SPSS

Opening SPSS and finding the correct menu can feel confusing if you are new to the software. Follow these exact steps to run your correlation:

1. Open your dataset in SPSS and make sure you are in the Data View.

2. Go to the top menu and click on Analyze.

3. Hover your mouse over Correlate and then click on Bivariate.

4. A new window will pop up. Find your first variable in the list on the left and click the arrow button to move it into the Variables box.

5. Do the same thing for your second variable so that both variables sit in the Variables box together.

6. Look at the Correlation Coefficients section in the middle of the window. Make sure the box next to Pearson is checked.

7. Look at the Test of Significance area. Choose either Two-tailed or One-tailed depending on your research hypothesis. Most researchers use Two-tailed.

8. Check the box that says Flag significant correlations. This puts little stars next to your important numbers.

9. Click the OK button to run the test and open your Output Viewer window.

Understanding the SPSS Output Table

Once you click OK, SPSS gives you a neat little grid called Correlations. Reading this table is simple once you know where to look. Let us break down the parts that matter.

The table shows rows and columns for each variable you picked. Where they cross, you see three pieces of information:

  • Pearson Correlation: This is your r value. It ranges from -1.0 to 1.0. A score of 0 means no relationship. A score close to 1 means a strong positive relationship. A score close to -1 means a strong negative relationship.
  • Sig. (2-tailed): This is your p-value. It tells you if your result is statistically significant. If this number is smaller than .05, you have a significant result.
  • N: This is the total number of people or cases used in the test.

Remember that a strong correlation does not mean one thing causes the other. Correlation just means the variables share a relationship.

How to Report the Results in APA 7 Style

When you write your results section for a thesis or a journal article, you must follow APA 7 rules. You need to report the correlation coefficient, the degrees of freedom, and the exact p-value.

Here is a template you can use:

A Pearson correlation was run to determine the relationship between study hours and exam scores. There was a strong, positive correlation between the two variables, r(48) = .65, p < .001.

Notice how we found the degrees of freedom. You take your total sample size (N) and subtract 2. If your N was 50, your degrees of freedom (df) is 48. Also, notice we did not write p = .000. SPSS often shows .000, but in APA style, you write p < .001.

Common Mistakes to Avoid

Students and researchers often make a few common errors when running this test. Keep these tips in mind:

  • Do not forget to check for outliers. One crazy data point can ruin your entire correlation score.
  • Do not claim causation. Just because study hours and test scores go up together does not mean studying caused the high grade by itself. Other hidden factors might be involved.
  • Do not skip checking your measurement levels. If SPSS thinks your scale variable is nominal, it will give you an error or a wrong test.

Frequently Asked Questions

What is a good correlation score?

There is no single magic number. In social sciences, an r value around .10 is small, .30 is medium, and .50 or higher is large. In physics or chemistry, researchers expect much higher numbers because their measurements have less natural noise.

Can I run a Pearson correlation with more than two variables?

Yes. You can throw five or ten variables into the Variables box at the same time. SPSS will spit out a giant grid that compares every single variable against all the other variables in a big matrix.

What do I do if my data is not normally distributed?

If your data fails normality checks, you should switch to a non-parametric test. Run Spearman’s rank-order correlation instead of Pearson.