Mediation Analysis

“Knowing that something works does not reveal the path through which it changed the outcome.”

01 / PLAINLY

What it means, plainly

Mediation asks whether part of an effect travels through an intermediate variable. That variable is called a mediator because it links a proposed cause to the outcome.

02 / CONTEXT

A little more

Mediation aims to separate direct and indirect effect pathways through an intermediate variable. A causal reading requires extra assumptions, especially about confounding of the mediator and outcome.

03 / WHY IT MATTERS

Why it matters

It supports questions about mechanisms while warning that a chain of associations alone does not prove the path.

04 / EXAMPLE

A familiar example

Tutoring might improve grades partly by increasing study time; study time is a proposed mediator.

05 / LIMIT

What it does not mean

A coefficient shrinking after adding a mediator does not by itself prove a causal mechanism.

06 / NOTICE

Notice it in your day

Draw tutoring → study time → grade, then add a direct arrow from tutoring to grade. Mark the indirect path.

FINAL NOTE

The idea worth keeping

Treat mediation as a causal question with assumptions, not an automatic coefficient subtraction.

QUESTIONS / 02

Questions people still have

What is a mediator?

An intermediate variable through which part of a proposed effect may pass.

Does a smaller coefficient prove causal mediation?

No. A causal reading requires additional design and confounding assumptions.