How to Report Correlation Results in APA Style (7th Ed.)

You have run the correlation in SPSS, the output is on your screen, and now comes the part that trips up more students than the analysis itself: writing the result up properly. A correlation is one of the simplest statistics to report in APA style, but examiners notice the small details — the degrees of freedom in parentheses, the missing zero before the decimal point, and the difference between a precise sentence such as “r(48) = .62, p < .001” and a vague claim that “the two variables were related.” This guide walks you through the whole process: running Pearson’s r and Spearman’s rho in SPSS, reading the output, and turning the numbers into a clean APA-style sentence and table that will survive a supervisor’s red pen.

What APA Style (7th Ed.) Requires for a Correlation

APA style asks for four pieces of information every time you report a correlation: the type of correlation coefficient, the degrees of freedom, the value of the coefficient, and the p-value. The standard format looks like this:

r(48) = .62, p < .001

Each part of that short string carries a job. The italic r tells the reader you used Pearson’s correlation. The number in parentheses is the degrees of freedom, which for a correlation is always the sample size minus two (N − 2). The value .62 is the correlation coefficient itself, and the p-value tells the reader whether the result is statistically significant.

Three formatting rules catch out almost everyone the first time. First, there is no zero before the decimal point for r or p, because a correlation can never be larger than 1 — so write .62, never 0.62. Second, report r to two decimal places and p to three decimal places. Third, never write p = .000. SPSS prints .000 when the p-value is very small, but APA requires you to write p < .001 instead, because a p-value is never exactly zero. Finally, report the sample size somewhere nearby — either in the text (“Fifty students completed the survey”) or in a table note — so the reader can check your degrees of freedom.

APA (7th ed.) also recommends reporting a confidence interval for the correlation alongside the point estimate, and an effect-size interpretation. For correlations, the coefficient itself is the effect size, and the usual benchmarks are .10 for a small effect, .30 for a medium effect, and .50 for a large effect. A sentence that says “a strong positive correlation” should be backed by a value of around .50 or above.

Step 1: Run a Pearson Correlation in SPSS

Before you can report anything, you need the output. If your variables are measured on an interval or ratio scale and the relationship between them looks roughly linear, Pearson’s r is the right choice. Here is the click-by-click route in SPSS:

  1. Click Analyze > Correlate > Bivariate.
  2. Move your two variables — for example, StudyHours and ExamScore — into the Variables box.
  3. Make sure Pearson is ticked under Correlation Coefficients.
  4. Under Test of Significance, leave Two-tailed selected unless you stated a directional hypothesis before collecting the data.
  5. Tick Flag significant correlations so SPSS marks significant results with asterisks.
  6. Click OK.

If you prefer syntax, or you want a record of exactly what you ran, paste this into a syntax window and run it:

CORRELATIONS
  /VARIABLES=StudyHours ExamScore
  /PRINT=TWOTAIL NOSIG
  /MISSING=PAIRWISE.

The PAIRWISE option tells SPSS to use every available pair of scores, which is the usual choice when a few values are missing. If you would rather drop any case with a missing value on either variable, replace it with /MISSING=LISTWISE.

Step 2: Run Spearman’s Rho When Pearson Does Not Fit

Not every pair of variables qualifies for Pearson’s r. If your data are ordinal — such as Likert-scale ratings — or the relationship is monotonic but not linear, or a scatterplot shows the normality assumption is badly violated, use Spearman’s rho instead. The clicks are almost identical:

  1. Click Analyze > Correlate > Bivariate.
  2. Move your two variables into the Variables box.
  3. Untick Pearson and tick Spearman.
  4. Leave Two-tailed selected and click OK.

The syntax version is:

NONPAR CORR
  /VARIABLES=StudyHours ExamScore
  /PRINT=SPEARMAN TWOTAIL NOSIG
  /MISSING=PAIRWISE.

In APA text, Spearman’s rho is written as rs (r with a subscript s). Everything else about the reporting format — degrees of freedom, decimal places, p-value rules — works exactly the same as for Pearson’s r.

Step 3: Read the SPSS Output

SPSS prints a Correlations table with one row and one column per variable. Three numbers in that table matter for your write-up:

  • Pearson Correlation — the value of r itself. In our example, the correlation between weekly study hours and exam score is .62.
  • Sig. (2-tailed) — the p-value. SPSS may print .000, which you will report as p < .001.
  • N — the number of cases used. Here N = 50, so the degrees of freedom are 50 − 2 = 48.

Before you write anything, check the value makes sense. The coefficient should fall between −1 and +1, the sign should match the direction you see in a scatterplot, and the sample size should match the number of cases you expected. If N is much smaller than your total sample, missing values may have eaten into your analysis — check whether listwise or pairwise deletion is doing what you intended. Also glance at the scatterplot (Graphs > Chart Builder, or Graphs > Legacy Dialogs > Scatter/Dot) to confirm the relationship is roughly linear and not driven by one or two extreme outliers. A single outlier can inflate or deflate a correlation dramatically, and no reporting format will save a result built on one.

How to Write the APA Sentence

The basic template is:

r(degrees of freedom) = .xx, p = .xxx

Applied to the example above, the sentence becomes:

“There was a strong, positive correlation between weekly study hours and exam score, r(48) = .62, p < .001.”

Note what the sentence does: it names both variables, states the direction (positive) and the strength (strong), and then gives the statistics in parentheses at the end. The words and the numbers agree — a reader should be able to tell the direction from the words alone and the size from the numbers alone.

For a negative correlation, the same pattern applies: “There was a moderate, negative correlation between daily screen time and sleep quality, r(78) = −.38, p < .001.” Note the minus sign in front of the coefficient.

For a non-significant result, report the exact p-value and say so plainly: “There was no significant correlation between caffeine intake and reaction time, r(33) = .12, p = .483.” Do not write “there was no correlation.” A non-significant result means the data did not provide enough evidence to reject the null hypothesis — it does not prove the correlation is zero. If the p-value is .051, say the result was not significant; do not describe it as “marginally significant” or “approaching significance,” which APA discourages.

For Spearman’s rho, swap the symbol: “There was a moderate, positive correlation between customer satisfaction rating and likelihood to recommend, rs(96) = .44, p < .001.”

Reporting in APA Style (7th Ed.): Full Examples

A single correlation fits neatly into one sentence, but most results sections report several correlations at once. When that happens, report each one in the same format and keep the sentence structure parallel so the reader can compare them at a glance:

“Exam score was positively correlated with weekly study hours, r(48) = .62, p < .001, and with lecture attendance, r(48) = .41, p = .003, but not with hours of part-time work, r(48) = −.09, p = .521.”

If you have three or more correlations, APA recommends a table instead of a paragraph. A correlation matrix table shows every pairwise correlation in one place, with the diagonal left blank or filled with dashes, and asterisks marking significance (usually * p < .05, ** p < .01). Report the coefficients to two decimal places, drop the leading zero, and add a note beneath the table giving the sample size and the meaning of the asterisks. Keep the table to the variables your research question actually needs — a matrix of twelve variables that nobody reads is worse than a focused table of four. For a step-by-step guide to building the table itself, see Creating APA Style Correlation Tables in SPSS.

APA (7th ed.) also encourages confidence intervals. SPSS does not print a confidence interval for a correlation in the standard Bivariate dialog, but if your version of SPSS supports it, click Bootstrap in the Bivariate Correlations dialog, tick Perform bootstrap, and request 1,000 samples — the output then includes a bias-corrected 95% confidence interval. Report it after the coefficient and before the p-value: r(48) = .62, 95% CI [.41, .77], p < .001.

Finally, remember that a correlation never proves causation. Your reporting should describe the association, not explain it: write “higher study hours were associated with higher exam scores,” not “studying more caused higher exam scores.” If your discussion section wants to argue for a causal explanation, that argument needs a design that supports it — a correlation on its own does not.

Common Mistakes to Avoid

  • Writing p = .000. SPSS prints .000, but APA requires p < .001.
  • Adding a leading zero. Write r = .62 and p = .03, never 0.62 or 0.03.
  • Forgetting the degrees of freedom. r = .62 on its own is incomplete; the reader needs r(48) to know the sample size behind it.
  • Reporting r without a direction word. Say positive or negative in words — the sign alone is easy to miss.
  • Claiming causation. A correlation describes an association; it cannot show that one variable caused the other.
  • Using Pearson’s r on ordinal data. Likert scales and ranked data need Spearman’s rho, not Pearson’s r. If you are unsure which one your data needs, see Pearson vs Spearman: Which Correlation Do You Need?
  • Hiding non-significant results. Report every correlation you tested, significant or not — selective reporting misleads the reader.

Key Takeaways

  • Report every correlation as r(df) = .xx, p = .xxx, with df = N − 2.
  • Drop the leading zero for r and p, use two decimals for r and three for p, and write p < .001 instead of p = .000.
  • Use Pearson’s r for interval or ratio data and Spearman’s rho (rs) for ordinal data or non-linear monotonic relationships. The full run-and-interpret walkthroughs are How to Run and Interpret a Pearson Correlation in SPSS and Spearman Rank Correlation in SPSS.
  • State the direction and strength in words, report non-significant results honestly, and never describe a correlation as proof of causation.
  • For three or more correlations, use an APA-style correlation matrix table instead of a long paragraph.

Frequently Asked Questions

Do I report the degrees of freedom for a correlation in APA style?

Yes. The degrees of freedom go in parentheses immediately after the r: r(48) = .62. For a correlation, df = N − 2, so with 50 participants you report r(48).

How do I report a non-significant correlation in APA style?

Give the exact p-value to three decimal places and state plainly that the result was not significant: r(33) = .12, p = .483. Do not write that there was “no correlation” — a non-significant result means the evidence was not strong enough to reject the null hypothesis.

How do I report Spearman’s rho in APA style?

Use the symbol rs in place of r, keeping the same format: rs(96) = .44, p < .001. State in words that you used Spearman’s rank correlation so the reader knows the coefficient is based on ranks, not raw scores.

Should I include a confidence interval when reporting a correlation?

APA (7th ed.) recommends it, but it is not required. If you include one, report it as 95% CI [.41, .77] after the coefficient and before the p-value. In SPSS, you can obtain a bootstrapped confidence interval from the Bootstrap option in the Bivariate Correlations dialog.

Need this done on your data? Get a free quote → /services/