How Good Is Your Memory, Really?

Memory keeps a strange filing system. It stores a commercial jingle from 2009 in triplicate, yet shreds the name of the person you met four minutes ago, along with any record of where the car is parked. Nobody fully controls which drawer things land in.

These seven questions are pure self-report, so honesty is the entire game. Grade yourself on names, errands, birthdays, and passwords, then claim your spot on the scale between goldfish and elephant. Fair warning: even the goldfish tier comes with perks the elephants quietly envy.

Result	Patterns	Share of patterns
Certified Goldfish	323	1.97%
The Squirrel	7,869	48.03%
The Owl	7,456	45.51%
Full Elephant	736	4.49%

Judging your memory and testing it are different jobs

Everything here is self-report, which makes it a study of your beliefs about your memory rather than of your memory. Researchers call that territory metamemory, and the general finding is humbling: how people rate their own everyday recall on questionnaires lines up only loosely with how they perform on objective tasks. Donald Broadbent's Cognitive Failures Questionnaire, published in 1982, is one of the long-running instruments in this area, and even well-built scales of that kind explain a modest share of measured performance.

Mood is a large part of the noise. People who are tired, low, or under strain rate their memories more harshly regardless of how they actually do, while a good week nudges the self-assessment upward. Ratings drift with how you feel about yourself this month.

Recall is reconstruction, not playback

The videotape metaphor is comforting and false. Frederic Bartlett showed as much in his 1932 book Remembering, where English readers retelling a Native American folk tale steadily reshaped it towards their own expectations, trimming the unfamiliar and adding tidy causal links that were never there. Elizabeth Loftus and John Palmer extended the point in 1974: viewers asked how fast cars were going when they smashed into each other gave higher speed estimates than those asked about cars that hit, and were later likelier to report broken glass that never existed.

Each retrieval is a rebuild from fragments, and confidence is a poor guide to accuracy. A vivid, detailed memory can be wrong in load-bearing places while feeling entirely solid.

Why names go missing first

Proper names are the weakest link in the chain because they are arbitrary. Researchers describe the puzzle through the baker and Baker pair: told a man is a baker, you can hang the fact on ovens, bread, and flour; told his surname is Baker, you have a sound attached to a face and nothing else. Names carry no meaning to catch on, so retrieval has fewer routes in, which is why one so often stalls on the tip of your tongue with the first letter tantalisingly available.

Taking the result in the right spirit

Nothing here is diagnostic. There are no norms behind the tiers, no controlled conditions, and no comparison with anybody else, so the outcome cannot say whether your recall is unusual for your age. Ordinary forgetting is extremely common and rarely means anything at all.

If you have noticed a real change, though, that is worth a proper conversation with a doctor rather than a quiz result. Clinical assessment uses validated batteries administered in person, and it exists precisely because self-judgement is unreliable in both directions.