Retention Experiment: Context-Stacking vs. Standard SRS

I've been testing a new method to stop forgetting words like 受領 and 乗り組み. By creating short 'action scenes' rather than single definitions, my 48-hour recall rate increased by nearly 50% in my last batch. It seems linking the word to a specific physical movement or object helps. Has anyone else measured the impact of active visualization on their retention stats?