Correction order biases rule interpretation

This proposed hypothesis asks why correction changes how people apply a rule in an invented story despite equal fact recall. It proposes that order leaves a belief about why an example was chosen. The test compares correction before retelling with retelling before correction. The example, source facts, example count and time remain identical; selection is explicitly random or triggered by a preceding error. Researchers then reset the selection policy and test how broadly the rule applies in new, held-out story units. If errors switch direction between general-rule and genuine-exception worlds, then fall below a specified criterion after a verified reset, that would support policy carryover beyond ordinary sampling-based inference.