How to Report Mediation Results in APA
Report the indirect effect (a × b) with a bootstrap confidence interval — the interval, not the Sobel z, decides significance — and report the direct and total effects alongside it.
What to report
For each mediation path, report:
- The indirect effect (a × b) with its bootstrap confidence interval (percentile or bias-corrected). Significance is judged by whether the interval excludes zero — not by a Sobel test.
- The direct effect (c′) of X on Y controlling for the mediator.
- The total effect (c) of X on Y.
- The path coefficients a (X → M) and b (M → Y).
- The number of bootstrap samples (commonly 5,000) and the interval type.
Avoid the old “full vs partial mediation” labels based only on whether the direct effect is significant; report the effects and let the indirect effect carry the inference. Most mediation is tested on cross-sectional data, which does not establish causal order — say so, and avoid causal wording (Maxwell & Cole, 2007).
APA template
We tested whether [mediator] mediated the relationship between [predictor] and [outcome] using [5,000] bootstrap samples. [Predictor] predicted [mediator] (a = [X.XX], SE = [X.XX], p = [.XXX]), and [mediator] predicted [outcome] controlling for [predictor] (b = [X.XX], SE = [X.XX], p = [.XXX]). The indirect effect was [X.XX], with a [95%] bias-corrected bootstrap confidence interval of [[X.XX], [X.XX]], which [did/did not] include zero, indicating [a significant / no] indirect effect. The direct effect was [X.XX] (p = [.XXX]) and the total effect was [X.XX] (p = [.XXX]). Because the data are cross-sectional, these results describe associations consistent with mediation rather than established causal mediation.
Common mistakes
- Relying on the Sobel test instead of a bootstrap confidence interval — the Sobel test assumes the indirect effect is normally distributed, which it usually is not.
- Reporting only the indirect effect and omitting the direct and total effects.
- Claiming causal mediation from cross-sectional data (Maxwell & Cole, 2007).
- Using “full” vs “partial” mediation as the headline conclusion instead of the effect estimates and their intervals.
Draft it with your own numbers
In Explonia, the “Article description” turns your own a, b, indirect, direct and total effects plus bootstrap intervals into a first-draft APA paragraph. You edit it and keep it; you stay the author. Results also export to SPSS/.sav, R and AMOS, with an APA report (.docx).
Frequently asked questions
Should I use the Sobel test or bootstrapping for mediation?
Use bootstrapping. The Sobel test assumes the indirect effect follows a normal distribution, which it rarely does, so it has lower power and can mislead. Report the indirect effect with a percentile or bias-corrected bootstrap confidence interval (commonly from 5,000 samples) and judge significance by whether the interval excludes zero.
Can I claim my mediator causes the outcome?
Not from cross-sectional data. A significant indirect effect is consistent with mediation but does not establish the causal order, because the variables were measured at one time point (Maxwell & Cole, 2007). Use associational wording and note the design limitation.
Do I still report direct and total effects?
Yes. Report the indirect effect (with its bootstrap interval), the direct effect (c-prime), and the total effect (c). Together they let readers see how much of the X to Y relationship runs through the mediator.
How to cite this page
If this page helped your research, you can cite it (APA 7):
Szymkowiak, A. (2026). How to Report Mediation Results in APA. Explonia. https://explonia.com/how-to-report/mediation-results-apa/
Last updated: July 2026
Run your whole study in one place
Build it, collect it, and get publication-grade analysis with a draft APA paragraph you can edit. Free to start. No credit card.
Start free