How Twenty-Four Students Helped Discredit Recovered Memories

In 1991, an undergraduate named Jim Coan asked his younger brother to read four short stories about the family's past. Three were true. The fourth, planted by Coan, described the brother getting lost in a shopping mall at age five and being rescued by an elderly stranger. After a few days of thinking about it, the brother began to fill in details — what the stranger looked like, how it felt to be alone — that nobody had given him.

Coan ran the experiment as an extra-credit project for his cognitive psychology professor, Elizabeth Loftus. Loftus and her student Jacqueline Pickrell formalized it in 1995 with twenty-four participants. About a quarter of them produced false memories of the mall trip, often elaborate ones. The paper landed in the middle of the recovered-memory wars then tearing through American family courts.

Here is what the study did not show. It did not prove that any specific person's memory of childhood abuse was implanted. It did not show that traumatic memories work like mall memories. The sample was tiny, the event was banal, the suggestion came from a trusted relative.

What the paper did show was an existence proof: under cooperative conditions, a meaningful fraction of adults will accept and embellish a false childhood event. That was enough. Defense attorneys started citing Loftus to challenge testimony from patients whose memories had surfaced in therapy. Some convictions were overturned.

A preregistered replication published in 2023 ran the protocol on 123 participants. Thirty-five percent produced a false memory — a higher rate than the original. The mall is a sturdier finding than its first sample suggested. What it actually means in a courtroom is still being argued.