Factorial analysis specification

Specify interactions contrasts and sensitivity

Derive the intended contrast from the fitted model rather than interpreting every coefficient as a marginal effect. Report combination outcomes and intervals for the planned effects and interactions.

Flow diagram

Factorial analysis specificationF1.S1Define factorial modelF1.S2Estimate prespecified contrastsF1.S3Apply multiplicity planF1.S4Assess adherence and missingnessF1.S5Report interaction uncertaintyF1.N1Factorial contrast planContrastModelEffect scaleMultiplicitySensitivityMain effect AMain effect BInteractionSimple effectsF1.N2Assumptions and rationaleF1.N3Method and protocol evidenceF1.S0 Start flowFactorial analysis specificationF1.S999 End flowEnd flow

Steps

Define factorial model

Specify outcome-appropriate model with factor indicators and planned interaction.

Estimate prespecified contrasts

Separate main, simple and interaction effects on their chosen scales.

Apply multiplicity plan

Use the planned claim structure and endpoint hierarchy.

Assess adherence and missingness

Handle factor-specific deviations consistently with estimands.

Report interaction uncertainty

Present interpretation limits and planned sensitivity analyses.

Factorial analysis specification

End flow

Tables and supporting information

Factorial contrast plan

Factorial contrast planContrastModelEffect scaleMultiplicitySensitivityMain effect AMain effect BInteractionSimple effects

Simple effects should be prespecified or clearly exploratory.

Assumptions and rationale

Interaction depends on scale. Avoid switching between pooled and stratified conclusions solely on a significance threshold.

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.

Sources