Posts Tagged ‘Julie Shaw’

Remembering Results that You Didn’t Find

May 8th, 2015 Comments off

Screen shot 2010-12-26 at 10.43.04 PMThere was a piece in the New Yorker in March called “Remembering a Crime that You Didn’t Commit.” It begins with a brief description of the McMartin Preschool case that omits virtually all of the evidence of guilt, while exaggerating the extent of satanic claims in the case. (See this post for more detail.)

The author then pivots to a description of Elizabeth Loftus’ Lost-in-the-Mall study, claiming that the experiment caused “six of the twenty-four test subjects [to] internalize the story, weaving in sensory and emotional details of their own.” But the vaunted fact checkers at the New Yorker were wrong about more than the McMartin case. They apparently failed to consult the original Lost-in-the-Mall study which makes it clear that the 25% figure includes those who “remembered” the event “either fully or partially,” and partially includes “remembering parts of the event and speculations about how and when it might have happened” (p.722). Speculations, of course, are not memories. And remembering, say, the “part” about being lost is hardly remarkable, since everyone has been lost at one time or another. In short, the original article never claimed that 25% of subjects internalized the false story. A much smaller number did.

Memory experts are quick to point out that memory is constructed and it changes over time. Curiously, those who cover memory researchers rarely, if ever, pause to ask whether professors describing their own research might reconstruct their “findings” over time, remembering results they didn’t actually find. That seems to be the real, but unrecognized, legacy of the Lost in the Mall study. To many, the study stands for much more than it ever actually found.

UPDATE (6/21/17): subsequent research has cast serious doubt on the other study, by Porter and Shaw, described in this article. Brewin and Andrews’ 2016 meta-study demonstrates that it is an extreme outlier. Pezdek and Blandon-Gitlin’s 2016 analysis concluded that the study was incoherent, employing “an unorthodox rating system” that rendered it “impossible to know what the high prevalence rate actually refers to.”