ECR-000001 Pre-Registration

Purpose

Run instrument calibration before broader hypothesis testing.

Hypotheses Tested

• H003 Multi-Model Convergence
• H012 Vocabulary Bias
• H013 Recognition Bias
• H015 Measurement Instrument Reliability

All remaining hypotheses are exploratory only.

Calibration Expectations

• Scientific Method should show richer iterative reasoning than Bloom.
• Scrum should differ from Scientific Method in execution-oriented ways.
• Models should return valid JSON if the schema is usable.
• Ambiguities should be recorded explicitly rather than being hidden by invented structure.