Session 13: Interaction

Question 1

Antecedent

Answer 1

A variable that comes before another variable. Note that all confounders are antecedents of the exposure they confound; all exposures are antecedents of the variables they mediate.

Question 2

Causal Partners

Answer 2

INUS causes in a sufficient cause are causal partners (also called, causal complements).

Question 3

Effect Measure Modification

Answer 3

When the association between an exposure and an outcome varies across categories or levels of another variable when using a particular measure of association (e.g. risk difference, risk ratio). This is visible in the data.

Question 4

Effect Modification

Answer 4

When the effect of an exposure of an outcome varies across categories or levels of another variable.

Question 5

External Validity

Answer 5

The extent to which the effect of an exposure found in a particular study extrapolates to another population or time period.

Question 6

Interaction

Answer 6

A synonym for effect modification. Effect modification is used more frequently for categorical variables; interaction is used more frequently for continuous variables.

Question 7

Redundancy

Answer 7

When an individual would complete two sufficient causes (i.e., causal pies) for the same outcome by the end of the study period.

Question 8

Synergy

Answer 8

One response type in regard to two exposures is a person who gets disease if and only if the two identified exposures are present. Such a person is a synergistic response type. This indicates that the two exposures are INUS causes in the same sufficient cause (i.e., they are causal partners).