A result such as p = .03 tells the reader a finding is statistically significant. It does not tell them how large or meaningful the effect is.
That is why stopping at the p-value is one of the most common ways SPSS-based dissertations lose marks. Here is how to report effect sizes properly.
The short answer
Always report an effect size alongside the p-value: Cohen’s d for two-group comparisons, eta-squared for ANOVA and odds ratios for logistic regression, ideally with confidence intervals and a plain-language interpretation.
Why a p-value is not enough
A p-value is influenced by sample size. With a very large sample, even a tiny effect can be statistically significant. With a small sample, a meaningful effect may fail to reach significance.
So a complete interpretation looks at the effect size alongside the p-value and, where appropriate, confidence intervals. The p-value answers is there evidence of an effect? The effect size answers how big is it?
The three effect sizes you will use most
| Effect size | Used for | Common benchmarks | How to read it |
|---|---|---|---|
| Cohen’s d | Comparing two group means (t-tests) | 0.20 small · 0.50 medium · 0.80 large | Difference in means relative to the spread of the data |
| Eta-squared (η²) | ANOVA | .01 small · .06 medium · .14 large | η² = .10 means about 10% of variance is associated with the factor |
| Odds ratio (OR) | Logistic regression and categorical outcomes | 1 = no difference in odds | Above 1 means higher odds, below 1 means lower odds than the reference group |
Treat benchmarks as rough guides. Whether an effect is practically important depends on your research context, not just on a cut-off.
Cohen’s d, eta-squared and odds ratios explained
Cohen’s d
Used when comparing two groups, such as treatment and control. It expresses the difference in means relative to the variability of the data.
Eta-squared
Shows the proportion of variance in the dependent variable associated with a factor in ANOVA. For more complex designs, partial eta-squared (η²p) is commonly reported.
Odds ratio
An interpretable measure of association for logistic regression. Always read it alongside its confidence interval and the research context.
What to report in your dissertation
A strong result reports the test statistic, p-value, effect size and, where appropriate, a confidence interval, followed by a clear explanation of what the finding means.
Reporting checklist
- Test statistic and degrees of freedom
- Exact p-value
- Effect size suited to the test
- Confidence interval where appropriate
- A plain-language interpretation of the finding
Your p-value shows whether the evidence is statistically significant, and your effect size shows how important the finding may be. Reporting both shows a more complete understanding of your results. Effect sizes matter just as much when a result is not significant, as our guide to reporting non-significant results explains.
Frequently asked questions
What is an effect size?
An effect size measures how large a difference or relationship is, rather than only whether it is statistically significant.
Does SPSS calculate Cohen’s d?
Recent SPSS versions report Cohen’s d with independent and paired t-tests. In older versions you may need to calculate it from the means and standard deviations.
What is a good eta-squared value?
As rough benchmarks, .01 is small, .06 medium and .14 large, but practical importance depends on your field.
How do I interpret an odds ratio?
An odds ratio above 1 means higher odds of the outcome and below 1 means lower odds, compared with the reference category.
Should I report effect sizes for non-significant results?
Yes. Effect sizes and confidence intervals help readers judge whether a non-significant result may still be meaningful.
Want your statistics chapter to score higher?
Analytivio’s statisticians can check your SPSS output, calculate the right effect sizes and help you report results clearly and correctly.
Keep reading
This article is general guidance. The right metrics and methods depend on your industry, data and goals.



2 Responses
[…] size may still deserve discussion, especially when statistical power is limited. Our guide to effect sizes explains how to report […]
[…] assuming the null hypothesis is true; it does not measure the size of your finding. Our guide to reporting effect sizes shows which measure to use for each […]