Factor definitions and estimands
Define main effects and interactions
Distinguish marginal effects averaged over the other factor, simple effects at a specific factor level and interactions. Name the outcome scale; absence of significant interaction is not evidence that effects are identical.
Flow diagram
Steps
Define factors and levels
Record intervention components, comparators and interactions of interest.
Define marginal and simple effects
Define marginal effects averaged over the other factor, simple effects at specified levels or interaction contrasts on a named outcome scale. Do not equate a nonsignificant interaction test with absence of interaction.
Define interaction questions
State the effect scale and clinically meaningful interaction contrast.
Specify endpoint priorities
Separate main, interaction, secondary and safety objectives.
Handle intercurrent events
Document factor-specific discontinuation, adherence and treatment-policy choices.
Factor definitions and estimands
End flow
Tables and supporting information
Factorial estimand register
Do not assume interaction is negligible merely to simplify the analysis.
Assumptions and rationale
Main effects average over other-factor assignments; interpretation can depend on interaction and allocation.
Method and protocol evidence
Attach the protocol, methodological sources and versioned software specification supporting these decisions. Record citations and provenance; do not enter participant identifiers.