FE-011A Analysis Plan

• Compare Version A vs Version B within each framework.
• Compare original-style vs redesigned versions across frameworks.
• Track whether redesigned versions improve evidence use, traceability, and actionability.
• Track whether redesigned versions increase complexity.
• Identify whether improvements are framework-specific or general.
• Do not claim success unless blinded evaluation favors redesigned outputs on multiple dimensions.
• Record cases where redesigns increase structure or verbosity without clear improvement in decision quality.
• Treat output length and section count as possible confounds, not automatic evidence of improvement.
• Report mixed or null findings explicitly rather than collapsing them into an overall redesign narrative.
• Avoid cross-framework claims unless the direction of effect is consistent and the tradeoffs are similar.
• Separate instrument findings from framework findings. A failure may indicate a flawed pilot design rather than a flawed framework redesign.