Walking Through a Door Erases What You Came For

Gabriel Radvansky and his Notre Dame lab published the doorway study in 2011 in the Quarterly Journal of Experimental Psychology. Across three experiments, college students picked up colored geometric solids at one table, carried them in a real shoebox or a virtual backpack, and dropped them at another table. Halfway between the tables they sometimes crossed a doorway. They were quizzed mid-trip on what they were carrying.

They forgot more after the doorway. Same distance, same time, same task — only the threshold changed, and recall dropped. A follow-up showed the effect even when participants only imagined walking through a doorway, which rules out anything to do with the physical movement itself.

Radvansky's explanation is that the brain runs short-lived "event models" — packets that hold whatever is locally relevant — and a doorway is a cue to close one model and open the next. The contents of the old room get marked stale and pushed out of fast access. It is good housekeeping that occasionally costs you the reason you walked into the kitchen.

Then the replication crisis came for it. A 2021 paper in BMC Psychology by Oliver Baumann's group at Bond University, working with collaborators including UCL's Daniel Yon, ran the test in three modalities — immersive VR, passive video, and real navigation — and found nothing. People barely forgot anything in any condition. The effect only reappeared when they loaded participants down with a concurrent counting task, hard enough to actually tax working memory.

That is the version that survives. The doorway is not magical; an idle, unburdened brain holds onto its plans through any number of thresholds. But under load — distracted, multitasking, walking and thinking — the brain genuinely does dump the room you just left, and it does it at the door.