10.1 Components of Writing Statistical Results
When writing up the results of a statistical test, we should usually include four major components:
- Description of your research question and/or hypotheses
- Description of your data
- Results of the inferential test
- Interpretation of the results
These four components mirror the logic of hypothesis testing: question → data → test → interpretation.
That does not mean every results paragraph has to follow the exact same sentence structure. There are many ways to write a clear APA-style results section. What matters is that the necessary information is included, the statistics are reported correctly, and the interpretation answers the research question without overclaiming.
Component 1: Describe the Research Question or Hypothesis
Start by reminding the reader what you were trying to find out. This does not need to be long, but the reader should understand the purpose of the analysis.
Useful sentence starters include:
- To examine whether [research question], I conducted [statistical test].
- I tested whether [independent variable/predictor] was related to [dependent variable/outcome].
- I tested whether [groups] differed on [dependent variable/outcome].
- I examined whether [predictor variables] predicted [outcome variable].
For example:
To examine whether students in Anastasia’s and Bernadette’s classes differed in grades, I conducted an independent samples t-test.
This sentence tells the reader the research question and the statistical test.
Component 2: Describe the Data
Next, describe the data in a way that helps the reader understand the result. This usually means reporting relevant descriptive statistics.
The exact descriptive statistics depend on the analysis:
| Analysis type | Descriptive information to report |
|---|---|
| t-test | Group means, standard deviations, and sample sizes |
| ANOVA | Group means, standard deviations or standard errors, and sample sizes |
| Correlation | Means and standard deviations for the variables, often with the correlation result |
| Regression | Descriptives and correlations are often reported in a table, especially when there are multiple predictors |
| Chi-square | Frequencies and percentages for the categories |
You do not need to report every descriptive statistic jamovi gives you. Report the statistics that help the reader understand the research question and the inferential result.
For example:
Anastasia’s students had higher grades (M = 74.53, SD = 9.00, n = 15) than Bernadette’s students (M = 69.06, SD = 5.77, n = 18).
This sentence describes the pattern in the data before reporting the inferential test. That is helpful because it tells the reader the direction of the difference.
If you are struggling to write this part, return to Chapter 5. Descriptive statistics are not separate from inferential statistics. They help readers understand what your inferential test is testing.
Component 3: Report the Inferential Test Results
The inferential test results are the formal statistical evidence. The exact statistics differ by test, but you will usually report:
- the statistical test used
- the test statistic
- degrees of freedom, when relevant
- the p-value
- the effect size
- a confidence interval, when relevant or available
For example:
t(31) = 2.12, p = .043, d = .74
Notice a few formatting details:
- Statistical symbols are usually italicized, such as t, p, M, SD, r, and d.
- There is no space between the test statistic and the degrees of freedom: t(31), not t (31).
- Exact p-values are usually reported to three decimal places.
- Report p < .001 when the value rounds to .000.
We will cover more of these details in Common APA-Style Errors.
Component 4: Interpret the Results
Finally, explain what the result means in plain language. This is where many students either stop too early or overclaim.
A good interpretation should:
- answer the research question
- describe the direction of the result, when relevant
- use reject/fail-to-reject language when discussing the null hypothesis
- avoid saying that a result proves something
- avoid saying that a non-significant result proves there is no effect
For a statistically significant result, you might write:
The result was statistically significant, so we rejected the null hypothesis. Anastasia’s students had higher grades than Bernadette’s students.
For a statistically non-significant result, you might write:
The result was not statistically significant, so we failed to reject the null hypothesis. The data did not provide sufficient evidence that grades differed between the two classes.
Notice that the second example does not say, “There was no difference.” A non-significant result means we did not find enough evidence to reject the null hypothesis. It does not prove that the groups are identical.
The p-value is important, but it is not the whole result. Your write-up should also include descriptive statistics, effect size when available, and a plain-language interpretation.
A Flexible Template
Here is a general template you can adapt:
To examine whether [research question], I conducted [statistical test]. [Describe the data using appropriate descriptive statistics]. The result was [statistically significant/not statistically significant], [test statistic and degrees of freedom] = [value], p = [value], [effect size] = [value]. [Plain-language interpretation].
This is a scaffold, not a script. Your final write-up should always fit the research question, statistical test, and audience.