Confounding, Mediation, Suppression, and Moderation

Post your commentary, answers, and any follow-up questions/confusions on Canvas the day before the lecture. (See the course website for suggestions regarding style and content.)

Readings

• Equivalence of the Mediation, Confounding and Suppression Effect
• Toward a better understanding of the influences on physical activity: The role of determinants, correlates, causal variables, mediators, moderators, and confounders

Access these readings as PDF, using the UW subscription for the corresponding publishers.

Reading questions

1. Explain confounding, mediation, suppression, and moderation. In your own words, explain each concept with an example and distinguish it from the other three.

2. Can confounding or mediation be established with statistical methods (i.e., is there a statistical test that shows whether a studied relationshiop between variables of interests is confounded or mediated)?

3. At a high-level, what statistical method do MacKinnon et al. describe to quantify the potential influence of a third variable on the relationship between two variables of interest?

4. Given a regression model of the form DV ~ IV (DV: dependent variable, IV: independent variable). Add a third variable (X) into the mix, and briefly describe two possible models – one for X being a confounder and one for X being a moderator.