How to report statistical results in APA: t-tests, correlations and regression

Your statistical output looks finished, but the moment you try to write it into your paper, every sentence starts sounding either too vague or too technical. You can see the t value, the correlation coefficient, or the regression table, but you are not sure which numbers belong in the paragraph, which ones belong in a table, and which interpretation counts as discussion rather than results. That uncertainty is exactly where students lose marks: not because they ran the wrong test, but because they reported the right test in a confusing way. If you are searching for how to report statistical results, you probably need more than a formula — you need a reporting pattern that works for APA style and still reads like academic writing.

Report statistical results in APA style by stating the test used, the main statistic, degrees of freedom where relevant, the p value, an effect size or model indicator, and a concise interpretation linked to the research question. The results section should tell the reader what the analysis found, not over-explain why it happened or what society should do about it.

In this guide

• How do you report statistical results in APA style?
• How do you decide which statistical results to report?
• How do you report t-test results in APA style?
• How do you report correlation results in APA style?
• How do you report regression results in APA style?
• How should tables and text work together when reporting statistics in APA?
• What mistakes do students commonly make when reporting statistical results?
• How can you revise a statistical results section before submission?

How do you report statistical results in APA style?

To report statistical results in APA style, write a complete sentence that identifies the test, gives the relevant statistic in APA format, reports the p value, and explains the direction or meaning of the finding. The sentence should be specific enough that another reader can understand the evidence without seeing your software output. Keep interpretation modest: the results section reports what the analysis indicates, while the discussion section explains why it may matter.

The basic APA reporting pattern

Most APA statistical reporting follows a simple structure:

1. Name the test and variables.
2. State the statistic and degrees of freedom if the test uses them.
3. Give the p value.
4. Add an effect size, confidence interval, or model fit statistic where relevant.
5. Interpret the result in relation to the research question.

For example, a psychology paper on exam anxiety might write: "Students in the high-anxiety group reported lower test confidence than students in the low-anxiety group, t(78) = -2.41, p = .018, d = 0.54." That sentence names the comparison, gives the test statistic, reports significance, includes effect size, and shows the direction of the difference.

Test statistic means the numerical value calculated by the statistical test, such as t, r, F, or β. p value means the probability of observing data at least this extreme if the null hypothesis were true, not the probability that your hypothesis is true.

What APA formatting looks like in results sentences

APA style uses italic letters for many statistical symbols: t, p, r, F, M, SD, R², and β. Report exact p values when possible, such as p = .032, rather than only writing p < .05. If the value is very small, use p < .001 rather than p = .000, because a p value is not literally zero.

Use two decimal places for many test statistics and means unless your department specifies otherwise. Correlations are often reported with two decimals, such as r = .42. p values are usually written to three decimals, with the leading zero omitted because p values cannot exceed 1.00.

How do you decide which statistical results to report?

Decide what to report by matching each statistic to your research question, hypothesis, and test choice. Do not paste every number from SPSS, Jamovi, R, Excel, or JASP into your paper. A good results section selects the statistics that answer the question and leaves software-only diagnostics or unused output out of the main narrative.

Start from the research question, not the output

Statistical reporting becomes easier when your variables and hypotheses are already clear. If your hypothesis asks whether two groups differ, a t-test result is central. If it asks whether two measured variables move together, a correlation is central. If it asks whether one or more predictors explain an outcome, regression results are central.

Students often run extra tests because the software makes them available. That usually creates a cluttered results section. Before writing, return to your variable definitions and measurement choices. If the variable logic is still unclear, reviewing how variable boxes connect to measurement indicators can help you decide which statistics actually belong in the paper.

Use a reporting map before drafting

A reporting map is a short plan that connects each research question to the exact result you will report. It prevents the common problem of writing a results section that lists numbers without answering anything.

Research aim	Weak reporting choice	Stronger reporting choice
Compare stress scores between first-year and final-year students	"There were some differences in stress."	"Independent-samples t-test comparing mean stress scores by year group, with M, SD, t, p, and d."
Examine medication adherence and age in a nursing sample	"Age was checked with adherence."	"Pearson correlation between age and adherence score, with r, p, and direction."
Predict customer satisfaction from service speed and staff courtesy	"Regression was significant."	"Multiple regression with overall model fit, predictor coefficients, confidence intervals, and explained variance."
Test whether online attendance relates to final quiz performance	"Attendance affected grades."	"Correlation or regression depending on the hypothesis, with careful wording that avoids causal claims."

Follow a small selection process

Use this sequence before writing your paragraph:

1. Write the research question in one sentence.
2. Identify the dependent or outcome variable.
3. Identify the grouping variable or predictor variable.
4. Confirm the test matches the measurement level and design.
5. Choose the statistic that directly answers the question.
6. Decide whether a table is needed for clarity.

If you are still choosing between tests, a statistical test decision structure for student research is a better starting point than trying to force the output into APA wording after the fact.

How do you report t-test results in APA style?

When reporting t-test APA results, state the type of t-test, the groups or conditions compared, the means and standard deviations, the t statistic, degrees of freedom, p value, and effect size. The sentence should make the direction of the difference clear. If the test is not significant, still report the statistic and avoid pretending that "no significance" proves no difference exists.

Independent-samples t-test

An independent-samples t-test compares the mean scores of two separate groups. Use it when each participant belongs to only one group, such as students who used flashcards versus students who used practice quizzes.

APA example:

"Students who used practice quizzes had higher exam preparation scores (M = 78.40, SD = 9.62) than students who used flashcards (M = 72.10, SD = 10.84), t(86) = 2.89, p = .005, d = 0.62."

This example gives the reader enough information to see both the statistical result and the practical direction. The effect size, Cohen’s d, helps show the size of the difference rather than only whether the p value crossed .05.

Paired-samples t-test

A paired-samples t-test compares two related scores from the same participants or matched pairs. Use it for pre-test/post-test designs, repeated measurements, or matched observations.

In a health sciences paper on a short patient education session, a student might write:

"Medication knowledge scores increased after the education session (M = 8.15, SD = 1.22) compared with before the session (M = 6.90, SD = 1.48), t(39) = 4.11, p < .001, d = 0.65."

This result does not prove that all education sessions work in all contexts. It reports that, in this sample and design, post-session scores were higher than pre-session scores.

Weak and stronger t-test reporting

Weak student version	Stronger APA-style rewrite
"The t-test showed that the new teaching method was better, and the result was significant."	"Students taught with the worked-example method had higher quiz scores (M = 81.30, SD = 7.94) than students taught with standard practice questions (M = 75.20, SD = 8.88), t(58) = 2.78, p = .007, d = 0.72."
"There was no difference between the two groups because p was over .05."	"The mean satisfaction score did not differ significantly between the online group (M = 3.82, SD = 0.71) and the in-person group (M = 3.95, SD = 0.68), t(64) = -0.75, p = .456, d = 0.19."

The stronger versions report the statistic, not just the verdict. They also avoid overclaiming. "Did not differ significantly" is safer than "there was no difference," because the study may not have had enough power to detect a small effect.

How do you report correlation results in APA style?

When reporting correlation APA results, state the correlation type, the two variables, the correlation coefficient, p value, sample size if useful, and the direction of the relationship. A correlation result describes association, not cause. Use wording such as "was positively associated with" or "was negatively correlated with" instead of "caused" or "led to."

Pearson and Spearman correlations

A Pearson correlation measures the linear association between two continuous variables. A Spearman correlation measures a rank-order association and is often used when assumptions for Pearson correlation are not met or when variables are ordinal.

A social sciences example might read:

"Social media comparison scores were positively correlated with body dissatisfaction scores, r(118) = .46, p < .001."

That sentence tells the reader that higher social media comparison scores tended to occur with higher body dissatisfaction scores in the sample. It does not say social media comparison caused body dissatisfaction. If your assignment requires effect interpretation, you can add a cautious sentence: "The association was moderate in size."

Reporting direction and strength

Correlation coefficients range from -1 to +1. A positive value means the variables tend to increase together. A negative value means higher values on one variable tend to occur with lower values on the other. A value near zero indicates little linear association in the sample.

Avoid vague phrases like "there was a relationship." Instead, name the variables and direction:

"Weekly study hours were positively associated with final assessment scores, r(92) = .31, p = .003."

For a non-significant result:

"Sleep duration was not significantly correlated with self-reported lecture engagement, r(74) = .12, p = .298."

Correlation does not equal prediction or causation

Correlation wording needs care because students often slip into causal language. In an education paper, "attendance improved grades" is stronger than the analysis supports if you only ran a correlation. A safer version is: "Attendance rate was positively correlated with final grade percentage."

If you later run regression, you may use prediction language, but still not causal language unless the design supports it. "Predicted" in regression means statistical prediction within the model, not proof that changing the predictor would cause the outcome to change.

How do you report regression results in APA style?

When reporting regression APA results, report the overall model first, then the individual predictors. Include R² or adjusted R², the model F statistic, degrees of freedom, p value, and relevant coefficients such as B, standard error, β, confidence interval, and predictor p values. Explain what the model suggests about the outcome without turning the results section into a discussion.

Simple linear regression

Simple linear regression estimates how one predictor is associated with one outcome. It gives a model-level result and a predictor-level result.

A business student studying service speed and customer satisfaction might write:

"Service speed significantly predicted customer satisfaction, F(1, 146) = 22.84, p < .001, R² = .14. Faster service was associated with higher satisfaction ratings, B = 0.38, SE = 0.08, β = .37, p < .001."

The first sentence reports the overall model. The second sentence reports the predictor. R² = .14 means the model explained 14% of the variance in customer satisfaction in the sample.

Multiple regression

Multiple regression estimates how several predictors relate to an outcome in the same model. This is common in undergraduate and master’s research papers because students often want to account for more than one variable.

A nursing example might examine whether age, discharge instruction clarity, and prior hospital admissions predict medication adherence after discharge:

"The regression model predicting medication adherence was significant, F(3, 102) = 9.47, p < .001, adjusted R² = .20. Discharge instruction clarity was a significant positive predictor of adherence, β = .39, p < .001, whereas age, β = -.08, p = .374, and prior admissions, β = -.12, p = .211, were not significant predictors."

This wording separates model fit from individual predictors. It also avoids saying that clear instructions "caused" adherence, which would require a stronger design than a standard observational regression.

What to include in a regression table

Regression paragraphs can become crowded. If your model has several predictors, use a table for coefficients and keep the paragraph focused on the model and the main predictors.

A typical regression table may include:

• Predictor name
• Unstandardised coefficient (B)
• Standard error
• Standardised coefficient (β)
• t value
• p value
• 95% confidence interval, if required

If your results section also needs descriptive statistics before inferential tests, a descriptive statistics table concept can help you decide what belongs before the regression output.

How should tables and text work together when reporting statistics in APA?

Tables and text should not repeat the same information word for word. Use the text to report the main finding and guide the reader; use the table to display detailed values that would interrupt the paragraph. APA-style reporting works best when the paragraph answers the research question and the table supports it.

Text for the main result, tables for detail

A good results paragraph gives the reader the answer first. For a t-test, that means the direction of the group difference and the test result. For a correlation, it means the direction and strength of association. For regression, it means whether the model was significant and which predictors mattered.

Tables are useful when you have multiple variables, several predictors, or a set of descriptive statistics. They reduce clutter and make comparisons easier. If the table already lists means and standard deviations, the paragraph does not need to repeat every value unless those values are central to the finding.

APA-style table habits

APA tables usually have clear titles, concise column headings, and no unnecessary vertical lines. They should be readable without requiring the reader to search through your paragraph for every abbreviation. Define abbreviations in a table note if needed.

For student papers, the exact table layout may depend on the assignment brief. Some instructors want APA tables embedded in the document; others accept simpler markdown-style tables in drafts. If your brief gives formatting rules, turn those requirements into a plan before writing. A structured approach to assignment brief requirements turning into a paper plan can prevent late-stage formatting fixes.

Results versus discussion

The results section should not explain the theoretical meaning of every finding. That belongs in the discussion. In results, write: "The relationship between workload and burnout was positive and significant." In discussion, you can connect that result to role strain theory, workplace policy, or previous studies.

This separation matters because examiners often mark structure as well as content. A paper that mixes results and interpretation can feel unfocused even when the statistics are correct.

What mistakes do students commonly make when reporting statistical results?

Students commonly make mistakes by reporting software output without interpretation, using causal language for non-causal tests, omitting effect sizes, misreporting p values, or mixing results with discussion. These errors make the analysis look less credible even when the test was appropriate. Most can be fixed by checking each sentence against the research question and APA reporting pattern.

Five common reporting mistakes

1. Writing a significance verdict without the statistic
   Student example: "The t-test was significant, so the hypothesis was accepted."
   Correction: Report the actual result: "The intervention group scored higher than the control group, t(52) = 2.34, p = .023, d = 0.64." Use "supported" rather than "accepted" unless your department specifically teaches that wording.

2. Using causal wording for correlation
   Student example: "More sleep caused students to achieve better grades, r = .29."
   Correction: Write association language: "Sleep duration was positively correlated with final grade percentage, r(88) = .29, p = .006."

3. Reporting p = .000 from software output
   Student example: "The regression was significant, p = .000."
   Correction: APA reporting uses "p < .001" when the value rounds to .000 in the software display.

4. Leaving out the direction of the finding
   Student example: "There was a significant difference between the groups."
   Correction: Name which group was higher or lower: "Participants in the guided reflection condition reported higher confidence than those in the standard instruction condition."

5. Putting interpretation into the results section
   Student example: "This proves that managers must increase flexible working because employees are happier."
   Correction: Keep the results sentence narrow: "Flexible working hours were positively associated with job satisfaction, r(104) = .34, p < .001." Discuss workplace implications later.

Why these mistakes matter

Statistical reporting is not only a formatting task. It shows whether you understand what your analysis can and cannot claim. A reader should be able to see the evidence, the direction of the finding, and the connection to the hypothesis without guessing.

Small wording changes often make the biggest difference. "Predicted," "correlated with," "differed from," and "was associated with" mean different things. Use the verb that matches the test and design.

How can you revise a statistical results section before submission?

Revise your statistical results section by checking accuracy, APA formatting, interpretation, and alignment with your research questions. Read each paragraph against the original hypothesis and confirm that every reported statistic has a purpose. A clean revision usually removes unnecessary output, adds missing effect sizes, and corrects overstatements.

A practical revision process

Use this sequence after drafting:

1. Match each paragraph to one research question or hypothesis.
2. Check that the test name matches the analysis you actually ran.
3. Verify every statistic against the original output.
4. Confirm that p values, degrees of freedom, and effect sizes are formatted consistently.
5. Remove software terms that do not belong in the paper, unless your instructor asked for them.
6. Replace causal verbs with association or prediction verbs where needed.
7. Move theoretical explanations into the discussion section.
8. Check that tables and paragraphs are not duplicating each other.

This process is especially useful in capstone and seminar papers where the results section sits between the methodology and discussion. If your methodology chapter is still changing, a clear methodology chapter stage from design to justification can help keep the analysis and reporting aligned.

Before you move on: APA statistical results checklist

• Each result is linked to a research question or hypothesis.
• The correct test name is used: t-test, correlation, simple regression, or multiple regression.
• Means and standard deviations are reported when they help interpret group differences.
• Test statistics are italicised where APA style requires it.
• Degrees of freedom are included for t-tests and model tests where relevant.
• Exact p values are reported unless p < .001 is appropriate.
• No result is reported as p = .000.
• Effect sizes or model fit statistics are included where expected.
• Correlation results are not written as causal claims.
• Regression findings separate overall model fit from individual predictors.
• Tables support the paragraph instead of repeating it line by line.
• Discussion-style explanations are saved for the discussion section.

Final wording check

Read your results section aloud and listen for three problems: unsupported certainty, missing direction, and unexplained numbers. "The result was significant" is rarely enough. "The intervention group scored higher than the comparison group" gives the reader the direction.

For undergraduate and master’s papers, clarity usually matters more than advanced statistical language. If your result answers the research question, follows APA style, and avoids overclaiming, it is doing its job.