Сравнение методов

	Робастный анализ модерации	Робастный анализ медиации
Область	Статистика	Статистика
Семейство	Latent structure	Latent structure
Год появления	2007	2008–2014
Автор метода	Hayes & Cai; Wilcox	Yuan & MacKinnon (median-regression formulation, 2014); robust bootstrap variants popularised by Hayes (2013) and Preacher & Hayes (2008)
Тип	Robust regression-based interaction test	Causal inference / indirect effects
Основополагающий источник	Hayes, A. F. & Cai, L. (2007). Using heteroscedasticity-consistent standard error estimators in OLS regression: An introduction and software implementation. Behavior Research Methods, 39(4), 709–722.	Yuan, Y., & MacKinnon, D. P. (2014). Robust mediation analysis based on median regression. Psychological Methods, 19(1), 1–20.
Другие названия	robust interaction analysis, robust moderated regression, HC-corrected moderation, outlier-resistant interaction testing	robust indirect effects, outlier-resistant mediation, robust causal mediation
Связанные	5	5
Сводка	Robust moderation analysis tests whether the effect of a predictor on an outcome depends on the level of a moderator variable, using estimation methods that remain valid under non-normality, heteroscedasticity, or the presence of influential outliers. It is the preferred approach when standard ordinary least squares assumptions cannot be trusted.	Robust mediation analysis estimates the indirect effect of an independent variable on an outcome through one or more mediators using estimators that resist the influence of outliers and non-normal error distributions. By combining robust regression (such as median or M-estimation) with percentile or bias-corrected bootstrap confidence intervals, it yields trustworthy conclusions when standard ordinary-least-squares mediation would be distorted by extreme observations.