How accurate are recovered memories?

Jun 17 2007

A decade or so ago, a spate of high profile legal cases arose in which people were accused, and often convicted, on the basis of recovered memories.

These memories, usually recollections of childhood abuse, arose years after the incident occurred and often during intensive psychotherapy.

So how accurate are recovered memories? The answer is not so clear. In fact, this question has lead to one of the most contentious issues in the fields of psychology and psychiatry.

Elke Geraerts, a postdoc of psychology at Harvard University and Maastricht University, the Netherlands, hoped to settle some of the controversy by enacting a large-scale research study examining the validity of such memories.

Recovered memories are inherently tricky to validate for several reasons, most notably because the people who hold them are thoroughly convinced of their authenticity. Therefore, to maneuver around this obstacle Geraerts and her colleagues attempted to corroborate the memories through outside sources.

The researchers recruited a sample of people who reported being sexually abused as children and divided them based on how they remembered the event. The memories were categorized as either spontaneously recovered (the participant had forgotten and then spontaneously recalled the abuse outside of therapy, without any prompting), recovered in therapy (the participant had recovered the abuse during therapy, prompted by suggestion) or continuous (the participant had always been able to recall the abuse).

Once all of the information was gathered, interviewers, who were blind to the type of abuse memory, queried other people who could confirm or refute the abuse events (other people who heard about the abuse soon after it occurred, other people who reported also having been abused by the same perpetrator, or people who admitted having committed the abuse him/herself).

The results, published in the July issue of Psychological Science, a journal of the Association for Psychological Science, showed that, overall, spontaneously recovered memories were corroborated about as often (37% of the time) as continuous memories (45%). Thus, abuse memories that are spontaneously recovered may indeed be just as accurate as memories that have persisted since the time the incident took place. Interestingly, memories that were recovered in therapy could not be corroborated at all.

Although the absence of confirmation that the abuse had happened does not imply that the memory is false, the findings of this study show that memories recovered in therapy should be viewed with a cautious eye, as the therapy context often involves an explicit effort to unearth forgotten memories and thereby raises the opportunity for suggestion.